Treatment of Stages IIIA and IIIB in Kienbock's Disease: A Systematic Review

Abstract

Background  Despite numerous proposed surgical interventions, there is a lack of consensus in the optimal treatment of advanced Kienbock's disease.

Purpose  This study aims to perform a systematic review of the current evidence in the management of Lichtman's stages IIIA and IIIB of the disease.

Methods  A literature search was performed using the MEDLINE, EMBASE, and COCHRANE databases to identify studies between 2008 and 2018 evaluating stage-specific outcomes in Lichtman's stages IIIA and IIIB. The quality of each included paper was evaluated using the Structured Effectiveness Quality Evaluation Scale (SEQES). Data extracted were stage-specificity, clinical and functional outcomes, and radiographic progression of Lichtman's staging.

Results  A total of 1,489 titles were identified. Eighty-three papers were fully reviewed, and 30 articles met eligibility criteria for inclusion. There were 3 low-quality and 27 moderate-quality papers. Surgical techniques reported included decompression surgeries, joint-levelling, and radial wedge osteotomies, revascularization techniques, intracarpal arthrodesis, proximal row carpectomy, arthroplasty, and balloon kyphoplasty. All treatment modalities offered pain relief and improvement in functional outcomes. Compared with proximal row carpectomy, intracarpal arthrodesis, and arthroplasty, nonsalvage procedures provided similar clinical and functional outcomes in both stages, with joint-levelling and radial wedge osteotomies preserving greater range of motion.

Conclusion  In this systematic review of Kienbock's disease stages IIIA and IIIB, all treatment modalities provided positive outcomes. In stage IIIB, there is evidence to support nonsalvage procedures, as they produced similar clinical outcomes to salvage procedures that have the advantage of not precluding future treatment options if needed and preserving greater range of motion.

Kienbock's disease is an uncommon disorder characterized by progressive avascular necrosis of the lunate, with eventual lunate collapse and pancarpal arthritis. ^{1 2} Establishing a treatment algorithm is challenging, owing to a paucity of comparative literature. This is particularly evident in Lichtman's stages IIIA and IIIB, where treatment recommendations are controversial and based on case series, nonrandomized comparative studies, underpowered, and uncontrolled studies. ²

As expected, without strong evidence to guide clinical decisions, surgeons have a varied approach to treatment. A survey among 375 members of the American Society for Surgery of the Hand in which 91% were practicing in North America, 4% in Europe, 3% in Asia, 2% in South America, and 1% in Australia revealed that radial shortening osteotomy and capitate shortening osteotomy were the most frequently performed procedures in stage IIIA, depending on the ulnar variance. However, respondents also included many other treatment modalities as their preferred choice, including nonoperative management with trial of splinting, vascularized bone graft (VBG), radial wedge osteotomy, radius (±ulna) core decompression, external fixation, pinning, intracarpal arthrodesis, proximal row carpectomy (PRC), and total wrist fusion. For stage IIIB, PRC was most frequently performed. Nonetheless, excluding external fixation and including lunate excision without replacement, surgeon preferences included the same procedures mentioned for stage IIIA with greater emphasis on PRC, intracarpal arthrodesis, radial shortening osteotomy, and total wrist fusion. ³ Similar disparities are seen among European surgeons, where radial shortening osteotomy is preferred in France and the United Kingdom for stage IIIB with negative ulnar variance, while scaphotrapeziotrapezoid (STT) arthrodesis, wrist denervation, radial shortening osteotomy, and PRC are performed in Germany with nearly equal frequency. ⁴

In light of the heterogeneity of treatment options and the general lack of consensus among hand surgeons, the purpose of this paper is to conduct a systematic review of the current evidence surrounding the treatment of Kienbock's disease stages IIIA and IIIB to compare clinical and functional outcomes, and radiographic progression for the various reported surgical techniques.

Methods

This systematic review was performed based on the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. A systematic search of all available literature was performed from the Medline, Embase, and Cochrane databases using keywords “Kienbock,” “Kienbock's disease,” “Kienbock disease,” “lunatomalacia,” “avascular necrosis of the lunate,” and “aseptic necrosis of the lunate.” English-language articles from 2008 to 2018 that incorporated treatment of Lichtman's stage IIIA and/or stage IIIB disease, reported objective clinical outcome measurements that were stage specific for IIIA and IIIB, had a minimum 1-year follow-up, and were within Sackett's level of evidence 1 to 4 were included for review. ⁵

Exclusion criteria were studies involving other hand/wrist diseases, skeletally immature patients, revision or secondary treatment, non-English papers, studies with fewer than three patients with stage IIIA and IIIB disease combined, Sackett's level of evidence 5 (expert opinion), case reports, and biomechanical studies. ^{5 6} Although there is no consensus on the minimum sample size for a case series, some authors have accepted case series with as low as three cases, while case reports typically do not exceed three patients. Therefore, a minimum of n  = 3 of stage IIIA and/or IIIB patients was chosen for this systematic review. ⁷

One reviewer (P.Q.W.) proceeded with the title and abstract screening to identify studies fulfilling the inclusion/exclusion criteria for full-text review. If the abstract was unavailable or uncertainty remained regarding the study's eligibility, the manuscript was then included for full-text review. Two independent reviewers (P.Q.W. and B.A.M.) then performed the full-text review followed by a structured effectiveness quality evaluation score (SEQES) assessment for each included article. A third reviewer (N.S.) was available in the event of any disagreement. This score evaluates the quality of a study's methodology with regard to seven general categories: (1) study question, (2) study design, (3) patients, (4) intervention, (5) outcomes, (6) analysis, and (7) recommendations. Each category contains subcategories that are given a score of 0 (criterion not met at all), 1 (criterion partially met), or 2 (criterion fully met). SEQES scores of 0 to 16, 17 to 32, and 33 to 48 denote poor, moderate, and high-quality studies, respectively ( Table 1 ). ⁶ For the SEQES evaluation, interrater reliability was assessed using an interclass correlation coefficient (ICC) average measurement, absolute-agreement, and two-way mixed effects model.

Table 1. Structured effectiveness quality evaluation scale (SEQES) criteria.

Study question
1. Was relevant background work cited to establish a foundation for the research questions?
Study design
2. Was a comparison group used?
3. Was patient status at more than one time point considered?
4. Was data collection performed prospectively?
5. Were patients randomized to groups?
6. Were patients blinded to the extent possible?
7. Were treatment providers randomized to the extent possible?
8. Was an independent evaluator used to administer the outcome measures?
Patients
9. Did sampling procedures minimize sample/selection biases?
10. Were inclusion/exclusion criteria defined?
11. Was an appropriate enrollment obtained?
12. Was appropriate retention/follow-up obtained?
Intervention
13. Was the intervention applied according to established principles?
14. Were biases due to the treatment provider minimized?
15. Was the intervention compared with the appropriate comparator?
Outcomes
16. Was an appropriate primary outcome defined?
17. Was an appropriate secondary outcome considered?
18. Was an appropriate follow-up period incorporated?
Analysis
19. Was an appropriate statistical test(s) performed to indicate differences related to the intervention?
20. Was it established that the study had significant power to identify treatment effects?
21. Was the size and significance of the effects reported?
22. Were missing data accounted for and considered in interpreting results?
23. Were clinical and practical significance considered in interpreting results?
Recommendations
24. Were the conclusions/clinical recommendations supported by the study objectives, analysis, and results?
Scoring methodology
0–criterion was not met
1–criterion was partially met
2–criterion was fully met
Total quality score
Low (0–16 points)
Moderate (17–32 points)
High (33–48 points)

Stage-specific outcome data for IIIA and IIIB were extracted. Outcomes of interest included pain scores, grip strength, range of motion (ROM), functional outcome scores, and radiological progression of Lichtman's staging. Where stage-specific outcome measurements were not reported but individual patient data were available, stage-specific averages were calculated for each possible outcome within a given study. Where possible, weighted averages were calculated for common outcome measures among studies with the same type of surgery.

Results

From the systematic search, 1,489 citations were identified, 522 were duplicates. Following title and abstract review, 83 articles met the criteria for full-text review. After full-text review, 30 were included in the systematic review. According to the SEQES evaluation, there were 3 low-quality and 27 moderate-quality studies with no high-quality papers ( Fig. 1 ). The ICC between raters was 1.0, indicating full interobserver agreement.

Fig. 1.

PRISMA and SEQES assessment flowchart. PRISMA, preferred reporting items for systematic reviews and meta-analyses; SEQES, structured effectiveness quality evaluation scale.

There was a total of 189 stage-IIIA patients and 197 stage IIIB patients included. Demographic data and study designs are summarized in Table 2 . Surgical interventions included decompression surgery, joint-levelling and radial wedge osteotomies, revascularization techniques, intracarpal arthrodesis, PRC, and arthroplasty. Other procedures described included distal capitate shortening osteotomy followed by capitometacarpal fusion to decompress the lunate ⁸ and a very distal radius wedge osteotomy where the aim is to increase the lunate covering ratio while preventing distal radial–ulnar joint (DRUJ) incongruity and subsequent DRUJ degenerative changes (a potential complication of shortening or traditional lateral closing wedge osteotomy). In this procedure, the apex of the closing wedge osteotomy is placed distally and is aimed toward the DRUJ in comparison to the traditional lateral closing wedge osteotomy. ⁹ Graner's procedure, an alternative to PRC but seldom-used technique, involves lunate excision, capitate osteotomy, and transposition of the proximal fragment to the radiocarpal joint to replace the lunate. The void is then filled with autologous bone graft. ¹⁰ Balloon kyphoplasty draws its inspiration from spine vertebroplasty and involves percutaneous injection of bone cement into the lunate. ¹¹ A novel technique used autologous bone marrow transfusion into the lunate, ultrasound therapy, and distraction using external fixation in an attempt to revascularize the lunate. ¹²

Table 2. Demographic data and study design.

Surgery type	Study (year)	Study design	Number of patients (male:female)		a Mean age (y)	b Mean F/u (y)
			IIIA	IIIB
Radius decompression	De Carli et al (2017) 49	Case series, retrospective	15 (11:4)	0	42	13
Lunate decompression	Fouly et al (2014) 8	Case series	4 (3:1)	0	25	1.7
Capitate shortening osteotomy	Singer et al (2017) 23	Case series, prospective	11 (NS)	0	30	3.0
	Afshar et al (2015) 21	Comparative, retrospective	9 (7:2)	0	29	3.1
	Gay et al (2009) 24	Case series, retrospective	6 (4:2)	0	41	5.6
Radial shortening osteotomy	Osada et al (2012) 25	Case series	4 (4:0)	6 (2:4)	45	2.2
	Afshar et al (2015) 21	Comparative, retrospective	12 (10:2)	0	33	3.2
	Mozaffarian et al (2012) 38	Case series, retrospective	0	24 (17:7)	33	4.6
	Calfee et al (2010) 34	Case series, retrospective	N/A	13 (9:4)	39	6.2
	Afshar et al (2013) 22	Comparative, retrospective	7 (4:3)	0	24	6.4
	Rodrigues-Pinto et al (2012) 30	Case series	12 c (NS)	6 (NS)	31	10.3
	Watanabe et al (2008) 26	Case series, retrospective	4 (2:2)	3 (3:0)	39	21
Lateral wedge ± shortening osteotomy	Shin et al (2017) 44	Case series, retrospective	14 (10:4)	N/A	30	7.1
Distal radial wedge osteotomy	Okubo et al (2017) 9	Case series, retrospective	1 (0:1)	5 (2:3)	49	2.7
Vascularized bone graft	Bürger et al (2014) 28	Case series	4 (3:1)	1 (1:0)	35	1.6
	Park et al (2016) 27	Case series	11 (6:5)	1 (0:1)	39	2.7
	Matsumoto et al (2018) 13	Case series, retrospective	16 (11:5)	10 (8:2)	35	5.2
	Afshar et al (2013) 22	Comparative, retrospective	4 (1:3)	0	22	6.5
	Fujiwara et al (2013) 31	Case series	10 (7:3)	See Below	44	12.3
VBG and shortening osteotomy	Fujiwara et al (2013) 31	Case series	See above	8 d (5:3)	44	12.3
VBG and capitate shortening osteotomy	Waitayawinyu et al (2008) 50	Case series, prospective	8 (NS)	0	25	3.4
Scaphocapitate arthrodesis	Özdemir et al (2017) 32	Case series, retrospective	0	9 (3:6)	33	2.3
	Luegmair and Saffar (2014) 36	Case series, retrospective	0	3 (NS)	35	8.8
STT arthrodesis	Hohendorff et al (2012) 14	Comparative, prospective	0	8 (8:0)	33	1
	Xu et al (2015) 33	Case series	0	24 (17:7)	38	1.0
	Lee et al (2012) 35	Comparative, retrospective	0	16 (10:6)	50	5.6
Proximal row carpectomy	Hohendorff et al (2012) 14	Comparative, prospective	0	11 (6:5)	33	1
	Buluç et al (2015) 29	Case series, retrospective	13 (4:9)	11 (4:7)	39	3.5
	Croog and Stern (2008) 19	Case series, retrospective	5 (NS)	12 (NS)	38	10
Graner's procedure	Facca et al (2013) 10	Case series, retrospective	1 (0:1)	2 (2:0)	37	25
Titanium lunate	Viljakka et al (2018) 37	Case series, retrospective	6 (6:0)	5 (5:0)	47	11
Pyrocarbon lunate	Visser et al (2016) 20	Case series, retrospective	0	15 (8:7)	38	2
Balloon kyphoplasty	Chen et al (2012) 11	Case series	2 (1:1)	1 (1:0)	35	2
Bone marrow, U/S, Ex-Fix	Ogawa et al (2013) 12	Case series, retrospective	11 (7:4)	3 (2:1)	44	5.3

Abbreviations: Ex-Fix, external fixation, F/u, follow-up, N/A, not applicable, NS, not specified in article, STT, scaphotrapeziotrapezoid, U/S, ultrasound, VBG, vascularized bone graft.

^aMean age is reported as the average age of all patients within the treatment group, regardless of disease stage.

^bMean F/u is reported as the average follow-up of patients within the treatment group, regardless of disease stage.

^cIncludes one stage-II patient.

^dFive patients underwent concomitant radial shortening osteotomy and two patients underwent concomitant capitate shortening osteotomy.

Clinical Outcomes

Mean values of many clinical outcomes (including functional outcome scores) were calculated using individual patient data points presented within studies. Pain results were either reported qualitatively (none, minimal, mild, moderate, and severe) or using visual analog scale (VAS) with a 0 to 10 scale. Studies using a VAS 0 to 100 scale were converted from 0 to 10 proportions. ^{13 14} ROM was inconsistently reported using the percentage of the contralateral side and/or in degrees of arc of motion. Grip strength was presented as a percentage of the contralateral side and/or in absolute strength in kilogram (kg) or Newton (N).

Functional outcome scores included the disabilities of the arm, shoulder, and hand (DASH), ¹⁵ quickDASH, ¹⁶ Mayo's wrist score/modified Mayo's wrist score, ^{8 17} Cooney's wrist function score, ¹⁸ Nakamura's scoring system for Kienbock (NSSK), ¹⁸ and patient-rated wrist evaluation/patient-rated hand and wrist evaluation score (PRWE/PRHWE) ^{19 20}

Stage-IIIA Results

There were 22 articles with stage IIIA results reported specifically in which 2 studies compared two different surgical interventions ^{21 22} ( Table 3 for clinical outcomes and Table 4 for functional outcome scores). All treatment modalities provided pain relief. Weighted mean VAS for patients who underwent capitate shortening osteotomy was 3.2 (three studies, n  = 26). ^{21 23 24} Despite heterogeneous pain reporting for radial shortening osteotomy, there were seven patients who were pain free, 19 had mild pain, and 1 had moderate pain at final follow-up (four studies, n  = 27). ^{21 22 25 26} VBG pain scores were also heterogeneous. Nonetheless, 11 patients were pain free at final follow-up, while only 3 patients had a VAS 4 or greater (three studies, n  = 31). ^{13 22 27} Pain relief was also achieved in VBG with medial femoral trochlea osteochondral flap reconstruction in four patients. ²⁸ Only 1 of 2 PRC studies presented stage-specific pain results. One patient was pain free while the other four had mild pain at an average of 10-year follow-up. ¹⁹

Table 3. Stage IIIA pain, range of motion, grip strength outcome measures.

Surgery	Study	n	Pain (VAS or qualitative)		Mean ROM (degree of Flex–Ext or % of contralateral side)		Mean grip strength(kg or % of contralateral side)
			Pre-op	Post-op	Pre-op	Post-op	Pre-op	Post-op
Radius decompression	De Carli et al (2017) 49	15	7	1.2	NS	77%	NS	80%
Lunate decompression	Fouly et al (2014) 8	4	Mod: 2 Severe: 2	None: 1 Mild: 3	80°	113.8°	NS	NS
Capitate shortening osteotomy	Singer et al (2017) 23	11	NS	4.2	NS	59% Flex 67% Ext	NS	60.6%
	Afshar et al (2015) 21	9	8.4	2.5	NS	66%	NS	29.3 kg 75.2%
	Gay et al (2009) 24	6	6.2	2.5	79.0° 62.0%	71.0° 54.7%	45.8%	70.8%
Radial shortening osteotomy	Osada et al (2012) 25	4	Mod: 4	None: 3 Mild: 1	88.3°	133.8°	54% (3/4)	77.8%
	Afshar et al (2015) 21	12	8.4	1.8	NS	68%	NS	28.3 kg 70.1%
	Afshar et al 2013 22	7	NS	None: 1 Mild: 5 Mod: 1	NS	88.7%	NS	33.1 kg 76.1%
	Rodrigues-Pinto et al (2012) 30	12 a	NS	NS	NS	NS	NS	79%
	Watanabe et al (2008) 26	4	NS	None: 3 Mild: 1	91.7°	103.3° (3/4)	ID	22.7 kg (3/4)
Lateral wedge ± shortening osteotomy	Shin et al (2017) 44	14	NS	1.1	NS	80.3%	NS	84.8%
Distal radial wedge osteotomy	Okubo et al (2017) 9	1	Daily pain: 1	None: 1	120°	170°	11 kg	33 kg
Vascularized bone graft	Bürger et al (2014) 28	4	NS	Relief: 4	80°	93.8°	NS	84.8%
	Park et al (2016) 27	11	7.2	2.5	79.8°	112°	59.5%	87.3%
	Matsumoto et al (2018) 13	16	Mild: 4 Mod: 9 Severe: 3	None: 8 Min: 4 Mild: 3 Severe: 1	71.3%	79.2%	59.3%	87.7%
	Afshar et al (2013) 22	4	NS	None: 2 Mild: 2	NS	76.5%	NS	32.5 kg 95%
	Fujiwara et al (2013) 31	10	NS	NS	66% Flex 73% Ext	78% Flex 78% Ext	57.9%	88.5%
VBG and capitate shortening osteotomy	Waitayawinyu et al (2008) 50	8	NS	NS	NS	71%	NS	72%
Proximal row carpectomy	Buluç et al (2015) 29	13	NS	NS	NS	77.7° 50.6%	NS	23.5 kg 65.7%
	Croog and Stern (2008 19	5	NS	None: 1 Mild: 4	90°	96° 69.4%	17 kg (4/5)	30.7 kg 83.8% (4/5)
Graner's procedure	Facca et al (2013) 10	1	NS	3	NS	55°	NS	35 kg
Titanium lunate	Viljakka et al (2018) 37	6	NS	0.3	82.5°	97.5°	33 b  kg	50 b  kg
Balloon kyphoplasty	Chen et al (2012) 11	2	7	0.5	72.5%	86.5%	42%	83%
Bone marrow, U/S, Ex-Fix	Ogawa et al (2013) 12	11	Mod: 6 Severe: 5	None: 8 Mild: 2 Mod: 1	79.1°	102.3°	80%	102.3%

Abbreviations: Ex-Fix, external fixation; Ext, extension; Flex, flexion; ID, incomplete data; Min, minimal; Mod, moderate; NS, not specified in article; Post-op, postoperative; Pre-op, preoperative; ROM, range of motion; U/S, ultrasound; VAS, visual analog scale; VBG, vascularized bone graft.

^aIncludes one stage-II patient.

^bConverted from Newton units, (fraction within brackets) = (number of patients with available data / sample size) when patient information missing.

Table 4. Stage-IIIA functional outcome scores.

Scoring system (bold)	Study	n	Mean score	Score (qualitative) A: excellent, B: good, C: fair, D: poor
Surgery				A	B	C	D
Mayo's wrist score
Radius decompression	De Carli et al (2017) 49	15	–	6	8	0	1
Lunate decompression	Fouly et al (2014) 8	4	85	–	–	–	–
Radial shortening osteotomy	Watanabe et al (2008) 26	4	77.7 (3/4)	–	–	–	–
Vascularized bone graft	Matsumoto et al (2018) 13	16	–	2	6	6	2
Vascularized bone graft	Fujiwara et al (2013) 31	10	–	5	4	1	0
Proximal row carpectomy	Buluç et al (2015) 29	13	67.3	–	–	–	–
Proximal row carpectomy	Croog and Stern (2008 19	5	77	–	–	–	–
Titanium lunate	Viljakka et al (2018) 37	6	68.3	–	–	–	–
Balloon kyphoplasty	Chen et al (2012) 11	2	77.5	–	–	–	–
Bone marrow, U/S, Ex-Fix	Ogawa et al (2013) 12	11	–	5	1	3	2
Disabilities of the arm, shoulder, and hand (DASH)
Capitate shortening osteotomy	Afshar et al (2015) 21	9	20.5	–	–	–	–
Radial shortening osteotomy	Afshar et al (2015) 21	12	24.2	–	–	–	–
Radial shortening osteotomy	Rodrigues-Pinto et al (2012) 30	12 a	8.7	–	–	–	–
Radial shortening osteotomy	Watanabe et al (2008) 26	4	6.8	–	–	–	–
Lateral wedge ± shortening osteotomy	Shin et al (2017) 44	14	14.2	–	–	–	–
Proximal row carpectomy	Buluç et al (2015) 29	13	12.4	–	–	–	–
Graner's procedure	Facca et al (2013) 10	1	27.3	–	–	–	–
Titanium lunate	Viljakka et al (2018) 37	6	4.7	–	–	–	–
Balloon kyphoplasty	Chen et al (2012) 11	2	12.1	–	–	–	–
QuickDASH
Capitate shortening osteotomy	Singer et al (2017) 23	11	44	–	–	–	–
Proximal row carpectomy	Croog and Stern (2008 19	5	17	–	–	–	–
Cooney's score
Capitate shortening osteotomy	Afshar et al (2015) 21	9	80.6	1	6	2	0
Radial shortening osteotomy	Afshar et al (2015) 21	12	80.4	1	9	2	0
Radial shortening osteotomy	Osada et al (2012) 25	4	82.5	–	–	–	–
Radial shortening osteotomy	Afshar et al (2013) 22	7	81.4	1	3	3	0
Vascularized bone graft	Afshar et al (2013) 22	4	88.8	3	1	0	0
Nakamura's scoring system for Kienbock (NSSK)
Capitate shortening osteotomy	Gay et al (2009) 24	6	–	0	2	1	3
Radial shortening osteotomy	Afshar et al (2013) 22	7	20.4	1	5	1	0
Radial shortening osteotomy	Rodrigues-Pinto et al (2012) 30	12 a	25.5	–	–	–	–
Lateral wedge ± shortening osteotomy	Shin et al (2017) 44	14	21.8	–	–	–	–
Distal radial wedge osteotomy	Okubo et al (2017) 9	1	–	0	1	0	0
Vascularized bone graft	Afshar et al (2013) 22	4	23.5	2	2	0	0
Bone marrow, U/S, Ex-Fix	Ogawa et al (2013) 12	11	–	3	6	2	0
Patient-rated wrist evaluation (PRWE)
Proximal row carpectomy	Croog and Stern (2008) 19	5	19.8	–	–	–	–

Abbreviations: Ex-Fix, external fixation; U/S, ultrasound.

^aIncludes one stage II patient, (fraction within brackets) = (number of patients with available data / sample size) when patient information missing.

Weighted average for ROM in percentage of the contralateral side were capitate shortening osteotomy: 61.5% (two studies, n  = 15) ^{21 24} ; radial shortening osteotomy: 75.6% (two studies, n  = 19) ^{21 22} ; VBG: 78.7% (two studies, n  = 20) ^{13 22} ; PRC: 55.8% (two studies, n  = 18). ^{19 29} One study showed a decrease in mean ROM postoperatively (capitate shortening osteotomy), ²⁴ 12 showed improvement, and 9 did not report preoperative and/or postoperative ROM.

Weighted mean grip strength as a percentage of the contralateral side was obtainable for most studies; capitate shortening osteotomy: 68.0% (three studies, n  = 26) ^{21 23 24} ; radial shortening osteotomy: 75.2% (four studies, n  = 35) ^{21 22 25 30} ; VBG: 88.2% (five studies, n  = 45) ^{13 22 27 28 31} ; PRC: 70.0% (two studies, n  = 17). ^{19 29} Ten studies showed grip strength improvement, while 11 did not report preoperative grip strength and 1 had incomplete data.

For functional outcome scores, weighted mean scores included Mayo's wrist score 70.0 (PRC, two studies, n  = 18), ^{19 29} DASH 15.1 (radial shortening osteotomy, three studies, n  = 28), ^{21 26 30} Cooney's score 81.1 (radial shortening osteotomy, three studies, n  = 23), ^{21 22 25} and NSSK score 23.6 (radial shortening osteotomy, two studies, n  = 19). ^{22 30}

Radiographic progression to Lichtman's stage IIIB was reported in one core decompression, one bone marrow transfusion with ultrasound therapy, four capitate shortening osteotomy, four radial shortening osteotomy, and four VBG patients. Stage-IV progression was reported in one core decompression patient. Three radial shortening osteotomy and three VBG patients improved to stage II at final follow-up.

Four stage IIIA treatment failures were reported. Two patients who had underwent capitate shortening osteotomy required revision surgery due to persistent pain. ²⁴ One patient who underwent PRC failed with subjective instability and weakness due to radial styloid, trapezium, and trapezoid abutment. The patient underwent subsequent radiocapitate arthrodesis with styloidectomy and was not included in the results of the original study. ¹⁹ One patient who underwent radial shortening osteotomy prior to Graner's procedure (thus not included in the pooled data), required subsequent wrist arthrodesis due to nonunion. ¹⁰

Stage-IIIB Results

Twenty-two articles with stage IIIB specific outcomes were identified, where one study compared two different interventions ¹⁴ ( Table 5 for clinical outcomes and Table 6 for functional outcome scores). Two of the five radial shortening osteotomy studies provided nominal pain scores. One study with short-term follow-up included four patients who were pain free and another two with mild pain. ²⁵ One long-term study reported three patients with only mild pain at an average of 18-year follow-up. ²⁶ Pain scores were heterogeneous for VBG. Nonetheless, three patients were pain free at final follow-up and no patient had a VAS score greater than 2.4 (two studies, n  = 11). ^{13 27} Intracarpal arthrodesis provided significant pain relief. Weighted mean VAS scores for scaphocapitate (SC) and STT arthrodesis combined were 6.2 preoperatively and 1.2 postoperatively (three studies, n  = 41). ^{14 32 33} One PRC study indicated six patients who were pain free, five who had mild pain, and one with moderate pain at an average of 10-year follow-up. ¹⁹

Table 5. Stage-IIIB pain, range of motion, and grip strength outcome measures.

Surgery	Study	n	Pain (VAS or qualitative)		Mean ROM (degree of Flex–Ext or % of contralateral side)		Mean grip strength (kg or % of contralateral side)
			Pre-op	Post-op	Pre-op	Post-op	Pre-op	Post-op
Radial shortening osteotomy	Osada et al (2012) 25	6	Mod: 6	None: 4 Mild: 2	96.7°	131.3°	53.5%	82.7%
	Mozaffarian et al (2012) 38	24	NS	NS	NS	84.8%	NS	81.1%
	Calfee et al (2010) 34	13	7.3	1.2	NS	102° 82%	NS	32.8 kg 77%
	Rodrigues-Pinto et al (2012) 30	6	NS	NS	NS	NS	NS	63%
	Watanabe et al (2008) 26	3	NS	Mild: 3	77.5° (2/3)	91.7° (3/3)	NS	30.7 kg
Distal radial wedge osteotomy	Okubo et al (2017) 9	5	Daily Pain: 5	None: 5	89°	120°	11 kg	24.4 kg
Vascularized bone graft	Bürger et al (2014) 28	1	NS	Relief: 1	100°	100°	NS	100%
	Park et al (2016) 27	1	7	2	40°	70°	62%	86%
	Matsumoto et al (2018) 13	10	None: 7 Mild: 3	None: 3 Min: 5 Mild: 2	66.1%	65.2%	59.5%	83.9%
VBG and shortening osteotomy	Fujiwara et al (2013) 31	8 a	NS	NS	68% Flex 75% Ext	80% Flex 89% Ext	53.4%	85.1%
Scaphocapitate arthrodesis	Özdemir et al (2017) 32	9	7.7	1.4	56.1°	68.3°	33.7%	71%
	Luegmair and Saffar (2014) 36	3	NS	NS	107°	87°	65%	69%
STT arthrodesis	Hohendorff, 2012 14	8	5.6	2.9	80.0° 54%	56.4° 39%	30.3 kg 52.9%	29.7 kg 62.1%
	Xu et al (2015) 33	24	5.9	0.5	NS	NS	18 kg	30.2 kg
	Lee et al (2012) 35	16	NS	NS	62°	68°	57%	83%
Proximal row carpectomy	Hohendorff et al (2012) 14	11	6	1.6	81.4° 62.5%	80.5° 57%	25.6 kg 38.6%	27.2 kg 69%
	Buluç et al (2015) 29	11	NS	NS	NS	83.2° 53.0%	NS	25.0 kg 70.0%
	Croog and Stern (2008 19	12	NS	None: 6 Mild: 5 Mod: 1	99.5° (11/12)	112.7° 83.5% (11/12)	22.3 kg (11/12)	36.0 kg 92.5%
Graner's procedure	Facca et al (2013) 10	2	NS	2.5	NS	62.5°	NS	50 kg
Titanium lunate	Viljakka et al (2018) 37	5	NS	0.86	88°	84°	26.6 b kg	41.4 b kg
Pyrocarbon lunate	Visser et al (2016) 20	15	4.8 (10/15)	2.7 (15/15)	95.4° (12/15)	72.3° (11/15)	23.5 kg (12/15)	29.0 kg (10/15)
Balloon kyphoplasty	Chen et al (2012) 11	1	8	2	58%	78%	40%	82%
Bone marrow, U/S, Ex-Fix	Ogawa et al (2013) 12	3	Mod: 2 Severe: 1	None: 1 Mild: 2	76.7°	93.3°	76.7%	93.3%

Abbreviations: Ex-Fix, external fixation; Ext, extension; Flex, flexion; Min, minimal; Mod, moderate; NS, not specified in article; Post-op, postoperative; Pre-op, preoperative; ROM, range of motion; STT, scaphotrapeziotrapezoid; U/S, ultrasound; VAS, visual analog scale; VBG, vascularized bone graft.

^aFive patients underwent concomitant radial shortening osteotomy and two patients underwent concomitant capitate shortening osteotomy.

^bConverted from Newton units, (fraction within brackets) = (number of patients with available data / sample size) when patient information missing.

Table 6. Stage-IIIB functional outcome scores.

Scoring system (bold)	Study	n	Mean score	Score (qualitative) A: excellent, B: good, C: fair, D: poor
Surgery				A	B	C	D
Mayo's wrist score
Radial shortening osteotomy	Mozaffarian et al (2012) 38	24	72.1	0	10	13	1
Radial shortening osteotomy	Calfee et al (2010) 34	13	84 (11/13)	–	–	–	–
Radial shortening osteotomy	Watanabe et al (2008) 26	3	72.7	–	–	–	–
Vascularized bone graft	Matsumoto et al (2018) 13	10	–	0	4	4	2
VBG and shortening osteotomy	Fujiwara et al (2013) 31	8 a	–	3	3	2	0
Scaphocapitate arthrodesis	Özdemir et al (2017) 32	9	–	0	5	4	0
STT arthrodesis	Xu et al (2015) 33	24	84.6	–	–	–	–
STT arthrodesis	Lee et al (2012) 35	16	71	4	10	0	2
STT arthrodesis	Hohendorff et al (2012) 14	8	57.9	–	–	–	–
Proximal row carpectomy	Hohendorff et al (2012) 14	11	66	–	–	–	–
Proximal row carpectomy	Buluç et al (2015) 29	11	67.3	–	–	–	–
Proximal row carpectomy	Croog and Stern (2008 19	12	88.3	–	–	–	–
Titanium lunate	Viljakka et al (2018) 37	5	67	–	–	–	–
Balloon kyphoplasty	Chen et al (2012) 11	1	75	–	–	–	–
Bone marrow, U/S, Ex-Fix	Ogawa et al (2013) 12	3	–	0	2	1	0
Disabilities of the arm, shoulder, and hand (DASH)
Radial shortening osteotomy	Rodrigues-Pinto et al (2012) 30	6	23.3	–	–	–	–
Radial shortening osteotomy	Watanabe et al (2008) 26	3	11.3	–	–	–	–
STT arthrodesis	Hohendorff et al (2012) 14	8	21.4	–	–	–	–
Proximal row carpectomy	Hohendorff et al (2012) 14	11	18.9	–	–	–	–
Proximal row carpectomy	Buluç et al (2015) 29	11	23.3	–	–	–	–
Graner's procedure	Facca et al (2013) 10	2	31.8	–	–	–	–
Titanium lunate	Viljakka et al (2018) 37	5	15.5	–	–	–	–
Balloon kyphoplasty	Chen et al (2012) 11	1	18.3	–	–	–	–
QuickDASH
Radial shortening osteotomy	Calfee et al (2010) 34	13	15	–	–	–	–
Proximal row carpectomy	Croog and Stern (2008 19	12	6.7	–	–	–	–
Cooney's score
Radial shortening osteotomy	Osada et al (2012) 25	6	85	–	–	–	–
Nakamura's scoring system for Kienbock (NSSK)
Distal radial wedge osteotomy	Okubo et al (2017) 9	5	–	0	5	0	0
Radial shortening osteotomy	Rodrigues-Pinto et al (2012) 30	6	22.5	–	–	–	–
Bone marrow, U/S, Ex-Fix	Ogawa et al (2013) 12	3	–	0	2	1	0
Patient-rated wrist evaluation/patient-rated hand and wrist evaluation
Proximal row carpectomy	Croog and Stern (2008 19	12	11.5	–	–	–	–
Pyrocarbon lunate	Visser et al (2016) 20	15	24.1 (14/15)	–	–	–	–

Abbreviations: Ex-Fix, external fixation; ROM, range of motion; STT, scaphotrapeziotrapezoid; U/S, ultrasound.

^aFive patients underwent concomitant radial shortening osteotomy and two patients underwent concomitant capitate shortening osteotomy, (fraction within brackets) = (number of patients with available data / sample size) when patient information missing.

Final ROM reported in degrees of flexion–extension arc of motion was obtained in 16 studies, whereas information as a percentage of the contralateral side was available in 8 articles. Weighted averages in degrees of flexion–extension arc of motion were radial shortening osteotomy: 108.6 degrees (three studies, n  = 22) ^{25 26 34} ; VBG: 85 degrees (two studies, n  = 2) ^{27 28} ; combined SC and STT arthrodesis: 67.1 degrees (four studies, n  = 36) ^{14 32 35 36} ; PRC: 92.1 degrees (three studies, n  = 33). ^{14 19 29} One study that included both PRC and STT arthrodesis showed a decrease in ROM postoperatively for both surgeries. ¹⁴ Four other studies (VBG, SC arthrodesis, titanium lunate arthroplasty, and lunate pyrocarbon implant arthroplasty) demonstrated a decrease in ROM compared with preoperative measures. ^{13 20 36 37} Otherwise, 10 studies reported improvement in ROM, 1 study showed no change, and 6 studies lacked preoperative and/or postoperative data.

Weighted mean grip strength as a percentage of the contralateral side with radial shortening osteotomy was 78.0% (four studies, n  = 49) ^{25 30 34 38} ; VBG: 85.4% (three studies, n  = 12) ^{13 27 28} ; combined SC and STT arthrodesis: 74.2% (four studies, n  = 36) ^{14 32 35 36} ; PRC: 77.6% (three studies, n  = 34). ^{14 19 29} One study showed a slight decrease in absolute grip strength compared with preoperative scores but an increase in relative grip strength when compared with the contralateral side in STT arthrodesis patients. ¹⁴ Seven studies did not report preoperative data.

Functional weighted mean scores included Mayo's wrist score 75.6 (radial shortening osteotomy, three studies, n  = 38) ^{26 34 38} ; Mayo's wrist score 75.6 (STT arthrodesis, three studies, n  = 48) ^{14 33 35} ; Mayo's wrist score 74.3 (PRC, three studies, n  = 34) ^{14 19 29} ; DASH 19.3 (radial shortening osteotomy, two studies, n  = 9) ^{26 30} ; DASH 21.1 (PRC, two studies, n  = 22). ^{14 29}

Radiographic improvement of Lichtman's staging to stage II was reported in one patient who underwent VBG. ²⁸ One patient who underwent radial shortening osteotomy progressed from stages IIIB to IV and eventually had a wrist arthrodesis performed due to persistent pain. ³⁴

Failures were reported in stage-IIIB patients. Two patients suffered from reflex sympathetic dystrophy post-PRC, presenting with persistent pain and edema. Both were treated conservatively, leading to poorer postoperative outcomes. ²⁹ One patient who underwent STT arthrodesis underwent subsequent removal of K-wires, complete denervation of the wrist, and radial styloid process resection. ¹⁴ Two titanium lunate arthroplasty implants dislocated. Both patients refused further treatment. ³⁷ In the lunate pyrocarbon implant arthroplasty series, one patient had a dislocation after 2 weeks, treated with closed reduction. Another patient was dissatisfied and underwent PRC 10 months after initial surgery. The latter patient was not included in the follow-up data in the original study. ²⁰

Discussion

Given the rarity of Kienbock's disease, most studies presented outcome data where multiple stages or diseases were combined together. Therefore, in this systematic review, clinical and functional outcome data were extracted, synthesized, and where possible, interstudy weighted means were quantified to present a comprehensive summary of the available literature on stage-specific treatments for stages IIIA and IIIB disease.

For stage IIIA, all surgical interventions provided good pain relief with most patients having none to mild pain at final follow-up whether short- or long-term. Relative ROM was generally higher in decompression procedures (excluding capitate shortening osteotomy alone), joint-levelling and radial wedge osteotomies, revascularization techniques, and balloon kyphoplasty compared with capitate shortening osteotomy alone and PRC. Revascularization surgeries, radial osteotomies, and decompressions resulted in greater relative grip strength.

There was an overall greater number of stage-IIIB patients who underwent salvage procedures, such as intracarpal arthrodesis, PRC, or arthroplasty, compared with radial osteotomies or revascularization surgeries. However, the latter group offered similar results. Radial shortening and wedge osteotomies, as well as revascularization, patients were either pain free or had mild pain at final follow-up, similar to intracarpal arthrodesis, PRC, arthroplasty, and balloon kyphoplasty. Joint-levelling and radial wedge osteotomies preserved the greatest arc of motion in flexion–extension, as expected as these are extra-articular approaches. ²² Relative grip strength was similar across all surgeries, though revascularization surgeries trended higher. The DASH and Mayo's wrist scores were comparable for radial shortening osteotomy, STT arthrodesis, PRC, titanium lunate arthroplasty, and balloon kyphoplasty. These results suggest that salvage procedures may be avoided in stage IIIB patients with similar expected outcomes. Considering that Kienbock's disease primarily affects the working adult and that nonsalvage procedures do not preclude future salvage procedures may be appealing to many patients. ³⁹

Nonsalvage procedures for late stages, especially IIIB, remain controversial. ^{1 3 4} One school of thought is to restore carpal height via intracarpal arthrodesis to prevent degenerative arthritis. ¹ Lunate implant arthroplasty or interposition arthroplasty remain an option to preserve motion, maintain carpus integrity, and keep future salvage surgery options available. However, complication rates are not minimal and little is known about long-term results. ^{20 37 40}

Despite satisfactory outcomes, PRC is not an ideal option for younger patients. Long-term consequences can include progressive degenerative changes and increased failure rate in younger and active patients. In a systematic review of long-term outcomes of PRC involving multiple wrist pathologies, including Kienbock's disease, radiographic findings were reported in five studies ( n  = 82) and demonstrated radiocapitate degenerative disease in 79.3% of patients. Where the radiocapitate arthrosis was graded as partial versus complete, 39.3% of patients had complete arthrosis. Meanwhile 35.1% of patients were graded severe arthrosis when classified as minimal, moderate, or severe. Additionally, 14.3% of patients (six studies, n  = 147) underwent revision to arthrodesis or arthroplasty at a mean of 53.4 months postoperatively due to failure. A significant proportion of failures were from patients under the age of 35, which could be a result of increased wrist loading. The authors recommended age and physical demands as a patient selection factor. ⁴¹

Meanwhile, in a study of 23 patients who underwent radial shortening osteotomy, including patients from stages I to IV, Nakamura et al showed that age (>30 years old) and the amount of shortening (>4 mm) were predictors of poor prognosis while preoperative staging and ulnar variance did not affect the results significantly. ⁴² In another study, where 41 patients (including stages II–IV) underwent either lateral closing wedge osteotomy or radial shortening osteotomy, a multiple regression analysis was performed to identify factors affecting clinical results. The authors found that gender (male), older age, and limited preoperative ROM were statistically significant negative predictors of clinical results. However, the preoperative Lichtman's stages and radiographic parameters were not a significant predictor of clinical results. ⁴³

While it is generally agreed that joint-levelling surgery may not halt further disease progression in late stages, many studies showed that it can provide satisfactory short- and long-term results while preserving motion and preserving eligibility for future salvage procedures. ^{1 26 44 45} It is thought that osteotomies or decompression surgeries around the distal radius help slow the rate of disease progression and inherently cause revascularization in the carpal area including the lunate during the postoperative hyperemic state given the anastomotic arterial network around the radius and carpus. ^{39 46} In a study of 24 patients (including stages I–IV), 9 were treated with radial shortening osteotomy, 10 with radial wedge osteotomy, and 5 nonoperatively. After undergoing serial magnetic resonance imaging (MRI), lunate T ₁ - and/or T ₂ -weighted signal intensity increased in all surgically treated patients, while none of the conservatively treated patients saw a signal intensity increase during follow-up MRI, suggesting an inherent revascularization process following osteotomies to the radius. ⁴⁷ Nonetheless, further studies are required to evaluate which surgical interventions, whether salvage or nonsalvage procedures, lead to better long-term outcomes, decrease the rate of reoperation, and identify risk factors affecting the decision making.

Limitations

There are several limitations to this study. The rarity of the disease inherently results in underpowered and uncontrolled studies, very few prospective comparative studies, and a significant number of studies without comparison of preoperative and postoperative values. Additionally, the objective of this study was to present stage-specific outcome data for Kienbock's IIIA and IIIB. However, since most studies did not report stage-specific outcomes, the number of included studies was limited. Combining data using weighted averages, especially with heterogeneous sample sizes, demographic data, and follow-up periods were the inherent source of bias.

Conclusion

The very low prevalence of the disease, unknown etiology, and scarce data often lead to poorly collected information in databases. The inconsistency of outcome reporting, small sample sizes, and lack of comparative trials increase the importance of high-quality studies to determine superiority of one treatment over another and to help guide the management of stages IIIA and IIIB of Kienbock's disease. While single centers may not have the volume to feasibly perform larger scale studies of this disease, collaborative multicenter efforts are needed. ⁴⁸ In order for these to be successful, researchers and wrist surgeons must agree on a unified set of core outcome measures. Ideally, this should be a disease-specific or site-specific outcome measure that is valid and responsive. While the NSSK offers a disease-specific score, it has yet to be validated. However, used in conjunction with wrist-specific scores (i.e., PRWE, Mayo's wrist score), this may be the best solution at present. The use of computer adaptive testing technology such as the patient-reported outcomes measurement information system (PROMIS) offers exciting possibilities in the future.