Basal joint arthritis of the thumb is a common site for arthritis in the hand and the prevalence is highest in women [1]. Different classifications have been described in the past [2,3]. The classification according to Dell combines radiological evaluation with clinical features and has our preference.

Basal joint arthritis of the thumb is also called arthritis of the carpometacarpal (CMC) joint of the thumb or first carpometacarpal joint (CMC I) or trapeziometacarpal (TM) osteoarthritis.

Most patients with trapeziometacarpal arthritis are treated conservatively. These treatments differ in efficacy and may be applied as single treatment or in combination with each other. Conservative options are drug therapy, splinting or bracing, injection therapy and physical therapy. If these therapies fail, or if the process of degeneration and arthritis continues, surgical intervention might be a next option.

Different surgical solutions are suggested in literature. The most common procedures are resection-arthroplasty, partial trapeziectomy, interposition arthroplasty [4], spacer interposition, fusion or arthrodesis [5] and prosthetic joint replacement [6]. These surgical interventions are not without the risk of the occurrence of a variety of complications. After any type of surgery for basal joint arthritis of the thumb complex regional pain syndrome (CRPS) type I may occur as a complication [7]. This is a debilitating acute pain syndrome, which could develop into a chronic disability. As there is no real cure for CRPS, one should emphasize on the prevention of what was formerly known as reflex sympathetic dystrophy (RSD). Our study group and others have reported a beneficial effect of the prophylactic vitamin C administration on the occurrence of CRPS in wrist fractures [8-10].

Now we will report our ongoing prospective results after a fully standardized treatment with a joint prosthesis for trapeziometacarpal arthritis under vitamin C prophylaxis.

We operated 40 hands in 34 patients. These 40 performed implantations for trapeziometacarpal arthritis were mainly under regional (or plexus) anesthesia and only 3 times under general anesthesia. Mean age of the operated patients was 60.8 years (SD = 8.1). There were 27 females and 7 males. In these 34 patients, there were 24 hands with stage III osteoarthritis according to Dell and 15 hands with stage II osteoarthritis according to Dell [2]. One patient had a traumatic dislocation of the trapeziometacarpal joint after a fall from the stairs. She was also treated with total joint replacement and with a plaster cast for 6 weeks postoperatively [16].

The median duration of complaints was 45.7 months (SD = 39.2).

The median time to surgery since the first complaints in patients with a Dell stage III was 48 months (interquartile range IQR = 36). For patients with a Dell stage II it was 24 months (IQR = 30). The p-value for the difference of time to surgery was 0.09.

First web opening increased with 15.4 degrees and there was a significant improvement for pain, ADL and satisfaction as well (p = 0.000). Patient satisfaction was strongly associated with the amount of pain reduction (Spearman’s rho 0.59) (Table 1).

Complications occurred in 6 of 40 surgical procedures (15%), but there were no infections.

In four cases a revision had to be performed (10%). Two times this was necessary because of a dislocation of the cup, and one because of a fracture of the trapezium.

A fourth revision was needed for an aseptic loosening of the cup after more than one year after the implantation. In these four cases a resection arthroplasty was carried out as a salvage procedure with fairly good results for the patients (due to the intention to treat policy, these results have been used in the statistical analysis).

Another patient suffered from a tenosynovitis of the thumb, which was treated successfully with a cortisone injection.

There was one patient with an entrapment of the dorsal sensory branch of the radial nerve that led to a temporary complex regional pain syndrome type II. This pain syndrome disappeared completely after surgical release of the nerve.

We detected no cases of complex regional pain syndrome type I according to the Veldman and IASP criteria.

One patient has recently died of natural course more than six years after her trapeziometacarpal joint replacement.

In a focused review by Shuler et al. different procedures for surgical intervention in basal joint arthritis of the thumb are extensively described resulting in different outcomes and complication rates [7]. What we can learn from this and the experience from others is that the complication rate for basal joint surgery of the thumb is high for all of these different procedures [4, 6, 17]. Normally the infection rate is low.

There is no evidence-based algorithm that gives us the best treatment for the patient’s complaints and the degree of osteoarthritis they suffer from. When we compare tendon position arthroplasty with CMC I arthrodesis, the more favorable procedure remains unclear. The range of motion is supposed to be better after interposition arthroplasty, but on the other hand pinch grip is stronger after arthrodesis [18]. The complication rate Mureau et al. report seems almost unacceptably high for the arthrodesis group (39%) and gives favor for the lower complication rate of 27% in the interposition arthroplasty group [19].

Persisting pain is another problem surgeons have to deal with. Either the procedure does not give enough primary pain relief or the resection or ligament reconstruction leads to mechanical painful abutment. Cooney et al. report satisfactory patient outcomes after second surgery with ligament reconstruction in case of failure of primary thumb trapeziometacarpal arthroplasty [20].

The revision rate after silicone implantation reported by Macdermid et al. was 23% in their series [6]. In our series the revision rate to resection arthroplasty was 10%.

Another complication is sensory loss on the dorsal aspect of the first ray. This may occur because the dorsal sensory branch of the radial nerve or the palmar cutaneous branch of the median nerve are at risk for peroperative traction or iatrogenic laceration, which could lead to a neuroma or complex regional pain syndrome type II (CRPS-II). Loss of sensation on the dorsal aspect of the first ray was reported in 14.5% by Saehle et al. which means in 8 out of 55 tendon interpositions [4].

Field et al. reported complex regional pain syndrome type I in 8% after ligament reconstruction and tendon interposition [21].

Takwale et al. reported even an incidence as high as 19% of CRPS-I in a special patient category with post-traumatic instability of the trapeziometacarpal joint [17].

In Table 2 we compare our outcome with Torrededia et al. who reported retrospectively about their results after 38 implantations in 34 patients with the same implant we used, the so-called Roseland prosthesis [22]. The difference with their postoperative treatment was that they splinted the first column for one month and started physical therapy one week after surgery. Their complication rate was 18.4%. CRPS type I occurred in 5 patients (13%). The length of the first column might in their opinion be the cause of CRPS in three of their 5 cases [22].

The differences with our study are the early functional treatment we encourage and of course the administration of the vitamin C. The difference in the prevalence of CRPS is significant though (Relative Risk (RR) = 0.87 (Confidence interval 0.77-0.98), p = 0.02).

Our patients have a satisfying pain relief and an improved and stable function. Patient satisfaction was strongly associated with the amount of pain reduction (Spearman’s rho 0.59). Patients with Dell stage III had logically a longer duration of symptoms before they were operated than patients with Dell stage II.

In general patients should be aware of the high complication and revision rate before they decide to undergo surgery. In case of failure of the procedure or the prosthesis, a salvage procedure can be performed. In such cases we prefer a resection arthroplasty.

The overall complication rate for trapeziometacarpal surgery is high and one of the possible complications is CRPS (variable incidence of 8 to 19 %) [17, 21]. Torrededia et al. described 5 cases of CRPS after 38 operations (13%) with the same implant we used, but under vitamin C we detected no CRPS in our study group [22].

Therefore we advise vitamin C as prophylaxis against CRPS in trapeziometacarpal joint replacement.