Osteoarthritis (OA) is the most significant cause of musculoskeletal morbidity in the elderly, and the hand is the most frequently affected area of the body.¹ A study in the United States found that among people older than 26 years, the prevalence of radiographic-confirmed osteoarthritis of the hand (HOA) and symptomatic HOA were 27.2% and 6.8%, respectively.²
The disability of HOA is primarily due to the combined impact of pain and decreased functionality, which significantly decreases quality of life for those who suffer from it.³ The small size of the bones and joints in the hand also make it a unique site for treatment. Current guidelines on the management of HOA are centered primarily on pain management, as there is a dearth of disease-modifying medical treatments, and surgical interventions are not as well developed as for OA of the knee, hip, and shoulder.
This paper aims to review the suspected causes and risk factors of HOA, to outline the appropriate diagnostic approach for a patient who presents with possible HOA, and to describe the appropriate management of a patient with HOA.
Evolving Etiology and Risk Factors
Once regarded as simply a “wear and tear” phenomenon, the etiology of OA has become remarkably complex. No longer thought of as solely a disease of cartilage, the model for OA development involves the entire joint and includes such pathological changes as loss of articular cartilage, osteophyte formation at the joint margins, subchondral bone remodeling with cysts and sclerosis, ligamentous dysfunction, muscle dysfunction, and synovial inflammation.⁴ Still, the basic biomechanical premise remains the same—accelerated damage to cartilage leads to bone-on-bone contact, pain, stiffness, and dysfunction. On the biochemical nature of the disease, some data posit that inflammatory cytokines play a role in altering the balance between cartilage buildup and breakdown,⁵ suggesting that a systemic understanding of OA is also necessary.
The pathogenesis of OA is dependent on the age at presentation. In the under-40 age group, for example, OA is almost always associated with some injury to the affected joint,² suggesting that in this minority of patients, trauma is a key factor in disease development. In patients older than 40, the pathogenesis appears not to rely on trauma, although it may be contributory. The most significant risk factor for HOA, or for OA of any joint, is age. The incidence of OA is greatly increased between the ages of 40 and 70, after which it seems to return to under-40 levels. The significance of the 30-year window of increased OA development is unknown, but it is useful in making a differential diagnosis. Any patient within the 40- to 70-year-old range or any patient with a history of joint trauma should be regarded as at risk.⁶
Although not necessary for the development of HOA, some kind of occupational or recreational history involving repeated hand motions is often present. This supposed link to the “wear and tear” hypothesis is simply that increased usage causes a thinning of the cartilage and when that damaged cartilage cannot be replaced, OA results. The effect of activity on HOA is dose dependent and shows differential distributions of joints involved depending on the particular repetitive task.⁷ Patients with a history of repetitive use damage should receive counseling on how to best care for their joints, and an immobilizing splint should be considered.
Whereas the connection between OA and obesity seems somewhat common sense in the context of the hip joint or the knee joint, obesity is also a known risk factor for the development of HOA,⁸ a finding that supports the systemic, biochemical nature of the disease. Proposed mechanisms focus on the endocrine function of adipose tissue possibly increasing the systemic levels of adipokines, such as leptin, and thereby impacting cartilage throughout the body.⁵ Published evidence from knee OA suggests that body fat percentage, not increased weight or body mass index (BMI), is the important feature in this relationship.⁹ Weight loss, when indicated, may be therapeutic in overweight patients with HOA and perhaps is preventative in those who have not yet developed the disease.
Female sex is a risk factor that is particularly predictive of OA of the hand.¹⁰ Unfortunately, another major cause of hand pain, rheumatoid arthritis (RA), also has higher prevalence in women. These two entities can be differentiated by a combination of age of onset, clinical presentation, joints presenting with pain, and laboratory testing. Family history is a pertinent risk factor,¹¹ although a specific genetic component has not been identified (see Table 1).
Diagnosis
Diagnosis of HOA requires a multifaceted approach, as there is no single test or finding that holds adequate sensitivity and specificity to make the diagnosis on its own (see Table 2). The European League Against Rheumatism (EULAR) published a set of recommendations for the diagnosis of HOA that were developed using a combined evidence-based and expert consensus methodology.¹² Prior to the EULAR task force, the best resource for the diagnostic criteria of HOA was the American College of Rheumatology (ACR) Criteria for the Classification and Reporting of Osteoarthritis of the Hand (1990).
The first step in diagnosis is comparing a patient’s complaints with the typical presenting symptomatology. HOA most often presents as hand pain on usage, with or without mild morning stiffness, affecting one or more joints of the hand. The distal interphalangeal joint (DIP) is most often involved, followed in order by the thumb base and proximal interphalangeal joint (PIP). This distribution of joints involved can be somewhat helpful in differentiating OA from other hand arthroses (Figure 1), but the specificity of this diagnostic approach is low.
Similarly, the presence of Heberden’s or Bouchard’s nodes may aid in the diagnosis but are not specific enough to be considered as a sole diagnostic marker. Some degree of functional impairment is to be expected but may be very similar to impairment seen in RA. An important subset of HOA is that of erosive OA. Patients with this condition may present with abrupt onset of marked pain, functional impairment, inflammatory signs and symptoms, and mildly elevated C-reactive protein (CRP). In general, these patients have a worse outcome and more rapidly developing symptoms than those with nonerosive OA.
Radiologic evaluation of HOA includes a posterior-anterior x-ray of both hands. Magnetic resonance imaging (MRI), scintography, and ultrasound have been evaluated as imaging modalities, but studies have yet to illustrate a true, reproducible benefit. X-ray indicators of HOA include joint space narrowing, osteophyte formation, subchondral bone sclerosis, and subchondral cysts. The correlation between x-ray findings and clinical OA is very weak in the hand, worse than in the hip or knee. There currently are no recommended interventions for asymptomatic HOA found incidentally on hand radiography.
There is no laboratory test for OA, but there are tests to evaluate alternative diagnoses. A strongly positive rheumatoid factor, for example, would lead one away from diagnosing OA. Inflammatory markers are not typically elevated, although CRP may be slightly above normal in erosive OA, as mentioned above. For a typical patient presenting with hand pain, the differential diagnoses include HOA, psoriatic arthritis, RA, and hemochromatosis.
Management of Pain
For the most part, managing HOA means managing the pain and dysfunction associated with HOA. To date, there is a lack of any drug or intervention that has been shown to significantly reverse the plethora of typical pathological changes seen in the course of the disease. Surgical interventions are available as a last-resort treatment in HOA of the base of the thumb but not at other affected joints of the hand. That said, because pain and dysfunction are by far the most common ailments in patients suffering from HOA, treatment of these two facets of the disease potentially provides vast improvements in patients’ quality of life, productivity, and satisfaction with medical care.
As with all pain management conditions, it is advisable to start with safer treatments moving on to more powerful options if safe options are ineffective (see Figure 2).
Non-drug Therapy: The EULAR strongly suggests that treatment of HOA be a combination of non-drug therapy and drug therapy. Non-drug therapies consist of patient education on how to avoid over-exerting the joints that are troubling them. Wrist and thumb braces have shown some effectiveness when combined with education and are attractive options because of their low risk profile. In one randomized trial, a full thumb-wrist brace was shown to be significantly more effective than a half brace.¹³
Acetaminophen: Acetaminophen is considered the first-line drug for OA of all joints because of its proven effectiveness, impressive safety record, and low cost. Although it has been used for decades to treat HOA, there are no randomized controlled trials investigating its effectiveness in this area. Extrapolated results from OA in other joints suggest that acetaminophen is less effective in decreasing pain and increasing function than nonsteroidal anti-inflammatory drugs (NSAIDs), but its safety record and cost are far superior, shifting the risk-reward balance in its favor. Using acetaminophen minimizes the risk of gastrointestinal (GI) side effects compared with NSAID use. Evidence strongly suggests that the often quoted hepatic damage with acetaminophen can be completely avoided simply by staying under the maximum daily dosage.¹⁴ Reports of renal damage, similarly, are unsupported by significant evidence.
Topical Treatments: Topical treatments have the advantage of safety. In particular, topical NSAIDs avoid the GI toxicity associated with oral formulations, and clinical trial data indicate that their effectiveness is roughly equivalent. Cost can be an issue with this mode of treatment, as most are still relatively new and are still unavailable in generic forms. If insurance plans allow them, topical NSAIDs are a reasonable second step in management if oral acetaminophen proves insufficient. Topical capsaicin, likewise, lacks systemic side effects and has proven effectiveness in treating pain. Topical treatment with both capsaicin and NSAIDs sometimes requires several days or weeks of application before full effectiveness is achieved.
Oral NSAIDs: Oral NSAIDs should be used as a third-line treatment after failure of both acetaminophen and topical treatments. Because of high GI toxicity, these drugs should be used only when absolutely required and for as short a course as necessary. This toxicity is dose dependent and increases with age, so oral NSAID therapy should be especially cautioned in elderly patients, and patient education should stress the importance of using as small a dose as possible. A proton pump inhibitor (PPI), misoprostol, or an H2 blocker can be added to protect the stomach lining. A cyclo-oxygenase-2 (COX-2) inhibitor also is an option in patients with GI risk, as studies have shown it to have gastric safety equivalent to that of NSAID plus a gastroprotective agent. Importantly, coxibs should be avoided in patients with significant cardiovascular (CV) risk factors, as they have been shown to enhance risk for cardiac events.¹⁵ Similarly, recent studies on non-selective NSAIDs have revealed an increased CV risk.¹⁶ Therefore, even non-selective NSAIDs should be used with caution in patients with CV risk factors, and the literature on this topic should be carefully followed for further updates on risk.
SYSADOA: The group of drugs now referred to as symptomatic slow-acting drugs for osteoarthritis (SYSADOA) includes glucosamine, chondroitin sulfate, avocado soybean unsaponifiables (ASU), diacerhein, and intra-articular hyaluronan.
A large meta-analysis of glucosamine, chondroitin, and their combination recently showed no impact on pain or structural OA in the hip and knee.¹⁷ Although this study did not investigate the effectiveness of these drugs in HOA, extrapolating the data is not encouraging. None of the other drugs in the SYSADOA category has been studied as rigorously as glucosamine and chondroitin. A systemic review on the effects of ASU on knee and hip OA suggest decreased pain and improved function in the treatment group compared with placebo.¹⁸ Diacerhein has been shown to have a small effect on decreasing pain and slowing progression in OA of the hip but also has been shown to be the only SYSADOA with significant side effects, as it causes diarrhea. There is inadequate evidence to suggest a large enough benefit to warrant the toleration of this side effect based on the evidence available today.
Intra-articular hyaluronan has been evaluated in the trapeziometacarpal joint in a study that found that 5 weekly injections decreased pain by 46% at 5 months post-treatment.¹⁹ Furthermore, a second trial indicates that hyaluronan injection may be more effective and longer lived than corticosteroid injection for this joint.²⁰
SYSADOA is a class of drugs with intriguing potential, as they have shown some very modest ability to modify the structural dysfunction in OA, but the evidence for their effectiveness in pain management is lacking, and their mechanism of action is not completely understood. The safety profile and modest findings suggest that the oral treatments in this group be left to patient discretion, as no side effects have been noted. Research on utility of hyaluronan injections should be followed, as its potential for replacing steroid injections is enticing. Finally, diacherin should not be recommended because of its lack of strong evidence and clear GI side effects.
Steroids: Expert consensus opinion suggests that intra-articular injection with long-acting corticosteroids (such as triamcinolone) is effective for flare-ups of HOA, especially in the trapeziometacarpal joint. No adequately powered randomized controlled trials have been published to support or refute the validity of this recommendation. Possible side effects from the injection of corticosteroids include localized depigmentation and weakening of the ligaments in the area, as well as known systemic side effects that go along with steroids.
Surgery: Although surgery remains a mainstay treatment for end-stage OA of the hip and knee, it is uncommonly used in HOA, and evidence for its effectiveness is lacking. However, for HOA at the base of the thumb, evidence does support effectiveness of surgical therapy when conventional therapies have failed. These surgical interventions include trapeziectomy, arthrodesis, osteotomy, ligament reconstruction, and joint replacement. Surgery for areas other than the thumb base is not yet widely available as a treatment option.
New treatments: This section very briefly reviews some possible new pharmacologic approaches to the management of OA yet to gain Food and Drug Administration approval.
Nerve growth factor (NGF) is a neurotrophin that has been identified as a major signaling molecule in the sensitization of nociceptors. Tanezumab, a monoclonal antibody against NGF, is currently under investigation in Phase III clinical trials and has been shown to significantly decrease OA pain when administered as an IV infusion every 8 weeks.²¹ Tanezumab also is under review for its effects on low back pain, pelvic pain, and neuropathic pain.
COX-inhibiting nitric oxide donors (CINODs) represent a potential solution to the GI and CV toxicity inherent in traditional NSAIDs. The structure of CINODs is that of an NSAID with a nitric oxide (NO) group covalently bound to it. The NO group is cleaved when the drugs enter the gut, thereby releasing the free NSAID to act as anti-inflammatory and analgesic, and the free NO moiety to counter the dangerous side effects of NSAIDS. In Phase III clinical trials, naproxcinod (the naproxen version of a CINOD) has shown non-inferiority to naproxen alone in treating knee OA, and early endoscopy studies show less GI damage.²¹ Furthermore, early data on CINODs demonstrate a decrease in blood pressure that may ameliorate the potential cardiovascular side effects inherent in COX inhibition.
Calcitonin is a hormone that participates in calcium and phosphorus metabolism. When given in pharmacologic doses, salmon calcitonin has been shown to affect osteoclasts, increase bone formation, inhibit cartilage-degrading metaloproteinases, and promote cartilage growth. An ongoing Phase III, randomized, placebo-controlled trial is investigating the effectiveness of salmon calcitonin in reversing some radiographic, structural changes in OA of the knee.²¹
Conclusion
Osteoarthritis in the hand is an incredibly common disorder in the elderly that can present in many different ways. Pain management in this setting should be individualized for specific patients based on the severity of symptoms, their desired functionality, and their individual risk factors. Treatment goals should be discussed and agreed on with the patient, as should steps for assessment of treatment effectiveness.