If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author: Ida Neergård Sletten, MD, PhD, Division of Orthopedic Surgery, Oslo University Hospital, Postboks 4950 Nydalen, 0424 Oslo, Norway.
To investigate the functional and aesthetic outcomes in a cohort with pollicizations performed due to congenital anomalies in our hospital.
Methods
From 1987 to 2016, we performed pollicizations in 32 hands of children aged 1 to 8 years (median, 2 years). We followed-up on 31 of the hands from 1 to 31 years (median, 10 years) after the procedure. The participants and their caregivers self-assessed their function and appearance with visual analogue scales and patient-reported outcome measures (Patient-Reported Outcomes Measurement Information System Pediatric Upper Extremity; the short version of the Disability of Arm, Shoulder and Hand Outcome Measure; and EQ-5D-3L). We examined the hands with regard to motion, strength, sensitivity, and function.
Results
There were 2 complications and 6 reoperations. Participants with mild anomalies (radial longitudinal deficiency Bayne type N/0 to 2) had better subjective and objective hand function than participants with severe anomalies (radial longitudinal deficiency Bayne type 3–4, ulnar dimelia, 5-finger hand). Hands with preoperatively near-normal index fingers had, in most cases, good thumb opposition and pinch, and hands in both groups benefited from the creation of a cylinder grip. Grip and pinch strength were lower than reported in cohort studies where an additional opponensplasty had been performed.
Conclusions
Hands with severe congenital anomalies also benefited from the procedure. We recommend a simplified follow-up program to identify cases where additional surgeries to enhance strength should be considered during growth of the child.
Children with congenital severe hypoplastic or aplastic thumbs lack cylinder grip to grasp large objects and rely on their ulnar fingers for pinch grip. Opponensplasty and ligament reconstruction are indicated when the thumb is moderately hypoplastic.
for congenital cases, consists of shortening and pronation of a neighboring finger (most often the index) and the creation of a thumb. Surgeons have, through the years, proposed modifications to the Buck-Gramcko technique.
Many hand units have performed cohort studies of pollicization outcomes since the 80s, and comparisons according to diagnosis have found the best results in cases with isolated thumb hypoplasia/aplasia.
Most of these studies were conducted in large tertiary referral centers, with low follow-up rates. Selection bias at study inclusion and attrition bias due to low follow-up rates can threaten a study’s validity.
Our hand unit is Norway’s largest, and we treat the majority of our nation’s patients with congenital upper limb anomalies (CULA). Norway is sparsely populated, and pollicization numbers per year are low. However, we have the ability to locate all former patients through their lifespan, as our hospitals’ systems are regularly updated according to our National Population Register. This registry contains social security numbers and residence information of everyone who resides or has resided in Norway. The purpose of this study was to investigate the surgical outcomes of all pollicizations performed in congenital cases, regardless of diagnosis in our unit.
Materials and Methods
Data collection, demographics, operative technique
The institutional review board of Oslo University Hospital approved the research protocol. We conducted the study according to the Helsinki declaration and reported it according to the Strengthening the Reporting of Observational Studies in Epidemiology Statement.
All participants >18 years and caregivers of participants <18 years signed a written consent form. We collected the participants from our unit’s surgical logbooks from 1984, through electronic searches for the International Statistical Classification of Diseases and Related Health Problems-9th revision procedure code “8471 Pollicization” between 1991 and 1999, and through the Nordic Medico-Statistical Committee procedure code “NDT69 Pollicization” between 1999 and 2016 in our hospital’s administrative system.
The first pollicization in a congenital pediatric case in Oslo was performed by Buck-Gramcko visiting our unit. Thereafter, we performed 32 pollicizations in 25 patients (13 boys and 12 girls) between 1987 and 2016. We invited all 25 patients to participate in the study, and 24 patients (31 hands; 97%) accepted the invitation and came for an outpatient follow-up evaluation. One patient declined to participate without giving a reason. We recorded the CULA type (Oberg-Manske-Tonkin classification, modified Blauth classification, and modified Bayne classification) and sorted the participants into 2 groups according to their diagnosis for stratified analyses of all outcomes (Table 1).
Group 1 (mild anomaly; n = 19) consisted of patients with radial longitudinal deficiency (RLD), modified Bayne type N/0 to 2. Group 2 (severe anomaly; n = 12) comprised patients with RLD, modified Bayne type 3 to 4, ulnar dimelia, or 5-finger hand (Fig. 1). Seven children had consecutive bilateral pollicizations, with a median interval of 10 months (4–17 months). Only 3 of 24 children had unilateral CULA as their sole congenital malformation. We performed an additional 1 to 6 upper limb surgeries (8 unilateral, 3 bilateral) in 14 of 24 children, most often ipsilateral radialization/centralization with or without predistraction in an external fixator (n = 7).
Table 1Descriptive Data for the 2 Study Groups According to Anomaly Severity
One bilateral pollicization in a case with bilateral 5-finger hand, not classified according to Blauth. One Townes-Brock syndrome with a 6-fingered hand with a Blauth IIIb thumb, second finger pollicized.
∗ Continuous data are given as medians and ranges.
† One bilateral pollicization in a case with bilateral 5-finger hand, not classified according to Blauth. One Townes-Brock syndrome with a 6-fingered hand with a Blauth IIIb thumb, second finger pollicized.
§ Case with ulnar longitudinal deficiency with a Blauth IV thumb, sorted into group 1 due to good preoperative index function.
Figure 1The anatomic differences between hands in the 2 anomaly severity groups, illustrated in two 20-year-old study participants. (A and B) The appearance and pinch function of 2 pollicized indexes at the ages of 3 and 4 years of a male participant in group 1 (RLD Bayne type 1, no wrist surgery). (C and D) The appearance and pinch function of the pollicized right index at the age of 1 year in a female participant in group 2 (RLD Bayne 3; radialization following soft tissue distraction at the age of 6 months). Not pollicized in left hand despite normal wrist and thumb aplasia due to many surgeries. Both were university students active in several sport activities.
Two surgeons (M.I.W. and Kjell Bye) performed 27 of 31 pollicizations, together or with another surgeon. Of the 4 remaining surgeries, 2 were the first 2 procedures performed by a surgeon nearing retirement and 2 were performed by a visiting surgeon. All used Buck-Gramcko’s technique. In most cases, we kept the second metacarpal base, secured the hyperflexed second metacarpal head to the base with sutures, and detached and reinserted all extensor and intrinsic tendons. Before this technique was standardized, we removed the base of the second metacarpal in 4 cases (the last in 2003) and used K-wires in 8 (the last in 1999). Microsurgical vein repair was necessary in 1 case due to iatrogenic laceration. We immobilized 25 thumbs in a cast, and 5 in a soft dressing for 1 month. One thumb had a cast for 8 weeks due to a release of tight radial soft tissue 4 weeks after the pollicization. The cast or soft bandage was followed by splinting for an additional month in 23 thumbs operated on between 1996 and 2016, while the first 8 thumbs started active mobilization at 1 month after surgery.
There were no other complications related to circulatory insufficiency or to bleeding, infection, marginal skin necrosis, or nerve injury.
One thumb malrotation required a first metacarpal rotational osteotomy at 4 years after surgery. Four additional thumbs had 5 secondary surgeries: 1 for release of soft tissue 1 month after surgery, followed by a corrective wedge osteotomy 1.5 years later; 1 for thumb flexor shortening after 1.5 years; 1 for extensor tenolysis after 2.5 years; and 1 for release of a palmar scar after 12 years.
At follow-up, all 15 children (18 thumbs) were full-time students and the 9 adults (13 thumbs) were either full-time students (n = 5) or had part-time (n = 1) or full-time (n = 3) occupations. The pollicized hand was nondominant in 17 of 18 hands where a unilateral pollicization had been performed, except for 1 case with bilateral ulnar longitudinal deficiency.
Subjective outcomes
A hand surgeon not involved in the treatment (I.N.S.) interviewed the participants and gave instructions for the patient-reported outcome measures (PROMs) and visual analogue scale (VAS) assessments. Caregivers for participants aged 5 to 17 years completed the parent proxy version of the Patient-Reported Outcomes Measurement Information System (PROMIS) Physical Function Upper Extremity Short Form 8 (UE 8), and participants aged 8 to 17 years completed the pediatric version of the PROMIS UE 8.
Participants aged >18 years completed the short version of Disability of Arm, Shoulder and Hand Outcome Measure for upper extremity function and the EQ-5D-3L for quality of life.
One caregiver for each of the 15 participants aged <18 years (18 hands), the 7 participants aged 8 to 17 years (9 hands), the 9 adult participants (13 hands), the hand surgeon (31 hands), and the occupational therapists (OTs; 31 hands) gave VAS assessments on how the pollicized digit “works like a thumb” and “looks like a thumb” between the extremes of “no” (0) and “yes” (100).
The adult participants (13 hands) and 1 caregiver for each of the participants aged <18 years (18 hands) gave VAS assessments between the extremes of “never” (0) and “always” (100) on “how often do you/does the child use the thumb to pinch versus scissor pinch for small objects” and “how often do you/does the child incorporate the thumb when holding larger objects like a bottle.”
Of the 7 participants aged 8 to 17 years, 4 were 9 to 11 years old and 3 were 12 to 14 years old. For this group, we regarded a VAS assessment of the 2 latter questions as too challenging.
Objective outcomes
Two pediatric OTs performed the measurements. They measured active range of motion (AROM) of the wrist, the first metacarpophalangeal (MCP) joint, and the interphalangeal joint with a goniometer (with 1-degree intervals). The Kapandji score and thumb retropulsion (cm) were recorded.
They graded active thumb palmar abduction and adduction along the radial side of the third finger as 1 to 5 (with 5 being the best) according to Zlotolow’s Likert scale pollicization pictogram.
We calculated the strength outcomes as percentages of population reference values according to sex, age, and side (left-sided reference values were chosen for nondominant hands).
They assessed the first carpometacarpal (CMC1) joint as stable or unstable, and used the Dellon Discriminator (2-point discrimination test, Fabrication Enterprises) and the Touch Test Sensory Evaluators (Semmes-Weinstein monofilaments, North Coast Medical) to assess pulp sensitivity.
The participants performed 3 acquisition tests (picking up a bead, a die, and a table tennis ball) and the sticker test. The latter was a modification of Zlotolow’s method, where a sticker is peeled from its backing.
The OTs detached a flexible toy sticker from its backing, and completely reattached it to a piece of paper placed on a table. The participants peeled off the sticker, and the OTs assessed their performances on all 4 tests as “able, easy”; “able, difficult”; “able by using ulnar fingers”; or “not able.”
Statistics
As the outcome data for most parameters were not normally distributed, we have presented continuous data as medians and ranges. We used the chi-square test and Fischer’s exact test for comparisons between the 2 severity groups for categorical outcomes and the Mann-Whitney test for continuous outcomes. For the objective outcomes, we had sample sizes of 19 hands in group 1 and 12 hands in group 2. We considered these sample sizes as large enough to claim a true difference if found statistically significant (alpha ≤ 0.05). For the subjective outcomes, we used different PROMs for participants under and over 18 years of age. In addition, there were participants aged <18 years that did not give VAS assessments of their hand function, and the participants aged >18 years were not accompanied by a caretaker. Therefore, the sample sizes for these outcomes became too small to perform statistical comparisons, and we have presented the raw data only.
Results
Subjective outcomes
Both groups had good PROM results for both upper limb function and life quality (Table 2). There were similar PROM results for participants with unilateral or bilateral CULA, regardless of severity group.
Continuous data are given as medians and ranges.PROMIS Ped UE 8 scores ranged from 10 to 57, with 57 being the best. PROMIS Parent Proxy UE 8 scores ranged from 13 to 55, with 55 being the best. QuickDASH scores ranged from 0 to 100, with 0 being the best. EQ-5D-3L index scores ranged from 0 to 1.00, with 1.00 being the best. EQ-5D-3L VAS scores ranged from 0 to 100, with 100 being the best. VAS scores ranged from 0-100, with 100 being the best.
Statistical comparisons were only performed where the subgroup sample sizes were regarded as sufficiently large (group 1, 19 hands; group 2, 12 hands).
PROMIS Ped UE 8
Participant > 8 y
40 (34–42)
35 (27–57)
PROMIS Parent Proxy UE 8
Caregiver
35 (24–45)
37 (28–55)
QuickDASH
Participant > 18 y
9 (7–25)
6 (5–39)
EQ-5D-3L index
Participant > 18 y
0.85 (0.62–1.00)
1.00 (0.62–1.00)
EQ-5D-3L VAS
Participant > 18 y
86 (50–95)
90 (85–100)
VAS: Works like a thumb
Participant > 8 y
90 (26–100)
47 (5–94)
Caregiver
71 (22–96)
37 (8–94)
Hand surgeon
94 (5–99)
51 (3–92)
<.05
OT
90 (46–100)
61 (1–77)
<.05
VAS: Looks like a thumb
Participant > 8 y
75 (46–96)
49 (14–100)
Caregiver
80 (4–95)
72 (22–94)
Hand surgeon
81 (8–97)
59 (24–88)
<.05
OT
72 (51–92)
58 (20–73)
<.05
VAS: How often do you/does the child use the thumb to pinch versus scissor pinch for small objects?
Participant > 18 y
100 (26–100)
46 (10–69)
Caregiver
88 (31–100)
22 (2–62)
VAS: How often do you/does the child incorporate the thumb when holding larger objects like a bottle?
Participant > 18 y
99 (3–100)
73 (63–100)
Caregiver
82 (42–100)
77 (18–94)
QuickDASH, short version of Disability of Arm, Shoulder and Hand Outcome Measure.
∗ Continuous data are given as medians and ranges.PROMIS Ped UE 8 scores ranged from 10 to 57, with 57 being the best. PROMIS Parent Proxy UE 8 scores ranged from 13 to 55, with 55 being the best. QuickDASH scores ranged from 0 to 100, with 0 being the best. EQ-5D-3L index scores ranged from 0 to 1.00, with 1.00 being the best. EQ-5D-3L VAS scores ranged from 0 to 100, with 100 being the best. VAS scores ranged from 0-100, with 100 being the best.
† Statistical comparisons were only performed where the subgroup sample sizes were regarded as sufficiently large (group 1, 19 hands; group 2, 12 hands).
For all assessors, the VAS ratings of overall function and appearance were better for group 1 (Table 2). The hands in group 2 used their thumbs less to pinch, but there were similar VAS results in the 2 groups regarding use of the thumb in the cylinder grip.
Objective outcomes
Extension and ulnar inclination of the wrist, MCP joint flexion, and thumb retropulsion were better in group 1 (Table 3). More patients had better Kapandji scores in group 1 (Table 3). Out of all 31 hands, 3 could not actively abduct their thumb palmarly beyond the third proximal interphalangeal (PIP) joint, all in group 2 (P < .05). Also, 7 of 31 hands could not actively adduct their thumb beyond the third distal interphalangeal (DIP) joint, 6 of these in group 2 (P < .05).
Missing data for the youngest participant (2 years) for radial and ulnar inclination and strength assessments, leaving 18 or 19 hands in group 1 and 12 hands in group 2, considered as sufficiently large sample sizes for statistical comparisons.
∗ Continuous data are given as medians and ranges.
† Missing data for the youngest participant (2 years) for radial and ulnar inclination and strength assessments, leaving 18 or 19 hands in group 1 and 12 hands in group 2, considered as sufficiently large sample sizes for statistical comparisons.
Strength was significantly better in group 1, and values were generally low (Table 3). Six hands had grip strength <10% of reference values, all in group 2. Almost all the hands that scored a 0 on the tip, lateral or palmar pinch (6 out of 8, 7 out of 8, and 6 out of 8, respectively), were in group 2.
Thumb length was measured between the third MCP joint and the PIP joint in 19 cases, between the PIP joint and DIP joint in 10 cases, and distal to the DIP joint in 1. There were 18 thumbs whose girth was less than or comparable to that of the fifth finger. There were 4 malrotated thumbs (2 at 30°–60° pronation, 2 with >120° pronation), 3 in group 2.
The CMC1 joint was unstable in 4 of 19 thumbs in group 1 and in 1 of 12 thumbs in group 2, but none of the cases were sufficiently severe to require additional surgery. Grip and pinch strength were similar for thumbs with stable (n = 25) and unstable (n = 5) CMC1 joints.
Pulp sensitivity was normal (green monofilament) in 26 of 30 thumbs, slightly reduced (blue monofilament) in 3 of 30, and reduced (red monofilament) in 1 of 30. Two-mm 2-point discrimination was present in 24 of 30 thumbs, while 3-mm 2-point discrimination was found in 6 of 30.
All 19 hands in group 1 picked up a bead and a die easily, but 4 of 12 hands in group 2 had to use ulnar fingers (P < .05). All hands in group 1 picked up a table tennis ball easily, whereas 3 in group 2 used ulnar fingers and 1 was unable to pick up the table tennis ball (P < .05). The sticker test was rated as “able, easy” for 14 of 19 hands in group 1 and 4 of 12 in group 2, as “able, difficult” for 5 of 19 hands in group 1 and 3 of 12 in group 2, as “able by the use of ulnar fingers” in 4 of 12 hands in group 2, and as “not able” in 1 of 12 hands in group 2 (P < .05). The 4 hands that used the ulnar fingers for the sticker test were the same that struggled with the pick-up tests, while the hand unable to peel off the sticker did not have problems with the pick-up tests. Of the 5 thumbs that had trouble with the functional tests, 1 had a Kapandji score of 6 and 4 had a score of 0. Their tip pinch was a median of 0% (0%–11%) of the reference value.
Discussion
Most of the outcomes in our study were in line with those of previous cohort studies on the subject. We confirmed earlier findings of better results for mild anomalies.
The thumbs with the poorest subjective and/or objective outcomes were almost exclusively in group 2, and no other patient predictor for a poor outcome was identified.
The similar PROM results in the 2 groups might be biased by a ceiling effect, small sample sizes, and bilateral CULA in two-thirds of the participants. The VAS assessments of function, appearance, and use of the pinch grip were better for participants with mild anomalies. This is in accordance with the findings of other studies, but differences in the questions asked and the choice of assessors make direct comparisons difficult.
The finding of similar cylinder grip VAS assessments might reflect that even though patients with severe malformations cannot expect as good a pinch grip as those with milder cases, they gain a functional cylinder grip.
Our participants’ thumb flexion, retropulsion, abduction, and adduction, as well as all strength outcomes, were significantly better in group 1, which was previously demonstrated in 4 studies.
This is explained by more severe tendon and joint hypoplasia in the complex malformations. AROM values were in accordance with these studies, but Kapandji scores and all strength values in both groups were lower in our study. The comparisons of strength are complicated by the use of different assessment methods and normal population references. The better results from de Kraker et al.
studies might be attributed to opposition transfers in 47% and 29% of their pollicizations, respectively. This was not performed in any of our patients. In a later study of Manske’s pollicizations, Goldfarb et al.
reported additional opponensplasty in 19 of 73 cases and later extensor tendon shortening or epiphyseal arthrodesis of the new thumb MCP joint in 15 of 73 cases. Based on our findings, we have become more aware of cases with suboptimal midterm results, and now consider additional procedures. We have performed the Huber opponensplasty in 2 of the study participants based on their follow-up results.
Thumb sensitivity, length, and rotation were satisfactory in almost all cases. The girth was less than in normal thumbs in all cases, as Goldfarb et al.
previously has demonstrated. We could not investigate whether the 5 unstable CMC1 joints were due to metacarpal (neo-trapezium) nonunion, as we did not obtain new radiographs at follow-up.
The sticker test gives an immediate impression of the thumb function in testing fine motor skills, and correlated well with the Kapandji score and the tip pinch strength. Comparison of the functional tests’ outcomes to previous papers was difficult, because there is no consensus on a pollicization score and many different activity-based functional tests have been used.
Researchers have recently developed video-based thumb hypoplasia evaluations that might be recommended for use in future studies, of which the Thumb Grasp and Pinch has the best documentation.
Also, in the hands with the most severe anomalies and the least favorable outcomes, we judged the index fingers preoperatively as suitable for pollicization. Spontaneous use of the index finger and satisfactory range of motion in the index PIP joint, combined with satisfactory positions and functions of the wrist and the elbow, have been necessary for meeting the indication for pollicization in our unit. Some of the cases are difficult to assess, and it would have been interesting to compare the functions of our pollicized hands with severe CULA with those that were never offered a pollicization. However, the latter group are likely to have even more severe anomalies, leading to selection bias in such a comparison.
The main strengths of our study were the high internal and external validities. The patient sample was representative for the population of interest without selection bias, as all operated cases in our unit were included. The same surgical team treated almost all the participants, and the technique was almost identical across cases despite a long inclusion period. We performed stratified analyses according to the 1 known prognostic factor for outcomes. There was a negligible attrition bias, as all except 1 thumb returned for the follow-up. The 4 studies with stratified analyses according to diagnosis that we used for comparison of our results were the studies with the highest quality we could find. Their follow-up rates varied between 44% and 76%.
The limitations of our study were the retrospective design and the relatively small number of pollicizations performed over a long time interval. The latter factor yielded differences in age and follow-up time that might be confounders, as strength, Kapandji scores, and dexterity scores improve with age in pollicized thumbs.
Only 6 thumbs were followed-up less than 5 years. When we performed stratified analyses according to the 2 malformation severity groups, the sample sizes were underpowered to perform statistical comparisons for many of the outcomes, including the PROMs and most of the VAS assessments, due to the participants’ age span. It would have been of some benefit to also include the 3 children aged >12 years in the VAS assessments of the pinch and the cylinder grip.
We believe it is valuable to demonstrate that patients’ outcomes after pollicization in smaller units in less densely populated countries are comparable for both subjective and objective parameters, with no more complications than in larger units. Almost 30 years ago, Manske et al.
also found that patients with severe anomalies gain some useful function from the procedure, mostly in the handling of large objects. Our study supports this finding through patient- and caregiver-reported VAS assessments of the cylinder grip. We changed practice after our study because we learned that additional surgeries (opponensplasty, extensor tendon shortening) to further enhance function should be considered at the time of the initial surgery and during the follow-up. To evaluate the indication for secondary surgery, we recommend a regular follow-up until the hand is fully grown. Measurements of grip strength, pinch, Kapandji score, retropulsion, and the sticker test are quick and valuable, and are recommended as a minimum test battery.
Acknowledgments
We thank occupational therapists Anne Birgit Stavenes and Inger Helen Bolstad for performing all physical measurements at follow-up. Thanks also to retired orthopedic and general surgeon Kjell Bye, who is one of the two surgeons who operated on most of the patients in this cohort.
References
Manske P.R.
McCarroll Jr., H.R.
Reconstruction of the congenitally deficient thumb.
00338-5&id=gr1.jpg){kind=link}