Pectoralis Minor Tenotomy with Occasional Secondary Neurolysis Significantly Reduces Self-Reported Pain and Headaches Across Heterogenous Chronic Pain Disorders of the Upper Limb

Reviewer 2

Overall Evaluation

This manuscript presents an ambitious and clinically relevant hypothesis—the Human Disharmony Loop (HDL)—supported by a relatively large retrospective case series (N = 318). The reported improvements in pain, range of motion, and symptom prevalence are substantial and suggest potential clinical value. However, the study design (retrospective, single-arm, single-surgeon) fundamentally limits causal inference. The manuscript currently overextends its conclusions, presenting mechanistic and unifying claims that are not supported by the methodological rigor of the study. While the work may have value as a hypothesis-generating contribution, significant revisions are required to improve methodological transparency, statistical validity, and alignment between results and interpretation.

We thank Reviewer 2 for the thorough and well-reasoned evaluation. We have undertaken substantial revisions to align the manuscript’s conclusions with the level of evidence supported by the study design. Specifically, we have (1) reframed the study explicitly as a retrospective observational case series, (2) softened mechanistic and unifying claims to hypothesis-generating language, (3) expanded the statistical analysis with paired tests, effect-size context, 95% CIs, multiple-comparison adjustment, and cross-diagnosis heterogeneity testing, (4) added subgroup pain-reduction analysis (Figure 5), (5) added a dedicated complications report, and (6) substantially expanded the limitations section. Specific line-referenced changes follow.

—

Major Revisions

Lines 20–22 (original).

The abstract frames HDL as a causal mechanism and implies that the intervention directly tests this hypothesis. However, the retrospective and non-controlled design does not allow causal inference. The language should be revised to reflect an exploratory or associative interpretation rather than a confirmatory or mechanistic one.

The abstract has been revised to use hypothesis-generating, associative language. The Background now uses “We hypothesize” rather than asserting a causal relationship, and the Conclusions describe HDL as a possible shared anatomic mechanism that “may constitute” a contributor rather than the cause.

Revised manuscript text: “In the Human Disharmony Loop (HDL), this neurologic asymmetry produces persistent protraction of the scapula… We hypothesize patients with the above who meet HDL diagnostic criteria will benefit from PM tenotomy with brachial plexus neurolysis (PM+ICN).”

Revised manuscript text: “The PM pathologizing the scapula may constitute a shared anatomic mechanism that contributes to chronic pain across heterogenous disorders of the upper limb.”

L21–26

L39–41

Lines 22–24 (original).

The HDL diagnostic criteria (medial coracoid tenderness and scapular protraction) are not validated, and no data are provided on reliability or reproducibility. This introduces substantial classification bias and weakens the internal validity of the study. The authors should justify these criteria or acknowledge their experimental nature more explicitly.

We have explicitly acknowledged that the HDL diagnostic criteria are not yet validated for accuracy or reproducibility. To partially mitigate classification bias in the current study, we describe that all exam findings (coracoid tenderness, scapular protraction, and SCT-positive lesions) were independently confirmed by a certified hand therapist (CHT) in addition to the surgeon. We acknowledge this is not equivalent to formal inter-rater reliability testing, which remains a needed step for future work.

Revised manuscript text: “All patients trialed at least 6 weeks of therapy before surgery and received an independent evaluation by a certified hand therapist (CHT) confirming all exam findings including coracoid tenderness, scapular protraction, and SCT+ neuropathic lesions.”

Revised manuscript text: “Diagnosis of HDL itself derives from history and physical, and this accuracy and reproducibility are not yet validated.”

L108–110

L270–271

Lines 23–25 (original).

The use of the scratch-collapse test for diagnosing neuropathy is problematic given its debated validity. The manuscript does not provide sufficient justification or acknowledge its limitations adequately, which undermines confidence in neuropathy-related outcomes.

We have explicitly acknowledged the limitations of the scratch-collapse test (with citation to Faszholz & Cheng 2024, a systematic review) and the absence of a gold standard for clinical neuropathy diagnosis (citing Gabriel et al. 2025). We have retained SCT as the primary neuropathy indicator because no validated alternative exists, but we now flag this limitation prominently in the limitations section.

Revised manuscript text: “Neuropathy was diagnosed via the SCT which has its limitations, although no gold standard exists.”

L271–272

Lines 93–97 (original).

The retrospective case series design, conducted by a single surgeon, introduces a high risk of selection bias, performance bias, and limited generalizability. The absence of multicenter data or independent validation further weakens external validity.

We have reframed the methods explicitly as a retrospective, single-center case series and have added a clear statement in the limitations section that the study lacks a control group, lacks independent validation, and lacks multicenter data. We now explicitly call for multi-institutional studies as the next step in establishing external validity.

Revised manuscript text: “This is a retrospective, single-center case series of consecutive patients presenting with chronic pain treated by a single fellowship-trained board-certified hand surgeon, between January 2023 to October 2025.”

Revised manuscript text: “Our study suffers numerous limitations. Chiefly, this is a retrospective study limited to one practice without a control group.”

Revised manuscript text: “The HDL should be seen as a proposed model which may contribute to unexplained chronic pain spanning various disorders, but ultimately needs substantiation via larger, multi-institutional studies that include a control group, utilize PROs, and incorporate objective testing to demonstrate mechanism of action.”

L98–100

L262–263

L283–286

Lines 98–100 (original).

The inclusion of a highly heterogeneous patient population (e.g., TOS, fibromyalgia, CRPS, cervical radiculopathy) without stratified or subgroup analysis makes interpretation of pooled results problematic. It is unclear whether the observed effects are consistent across conditions or driven by specific subgroups.

We have added a per-diagnosis subgroup analysis of pain reduction (Figure 5 forest plot, with 95% CIs by diagnostic category) and a formal cross-group heterogeneity test using Cochran’s Q and I² statistics on inverse-variance-weighted within-group means. The pain reductions were statistically indistinguishable across all eleven diagnoses (Cochran’s Q = 7.2, df = 10, p = 0.709, I² = 0.02%), demonstrating that the pooled effect is not driven by any single subgroup. We note this consistency is descriptive and does not in itself prove a unified mechanism.

Revised manuscript text: “The reductions in pain were statistically indistinguishable across all diagnoses (Cochran’s Q = 7.2, df = 10, p = 0.709, I² = 0.02%). (Figure 5)”

Revised manuscript text: “Figure 5. Pain Reductions by Chronic Pain Diagnosis.”

L154–156

L161

Lines 100–107 (original).

The intervention combines surgery (PM+ICN) with postoperative physiotherapy. Since these components are not separated analytically, it is not possible to attribute observed improvements specifically to the surgical procedure. This represents a major confounding factor.

We agree this is a major confounder. We have added a detailed description of the standardized PT protocol (timing, content, and duration) and explicitly acknowledged in the limitations section that the relative contributions of surgery and rehabilitation cannot be separated in the current design.

Revised manuscript text: “Each patient underwent PM+ICN via an open deltopectoral approach as previously described. This was followed by a standardized PT protocol consisting of brachial plexus and axillary nerve glides at week 2 and medial rhomboid and upper trapezius strengthening with scapula retraction postural taping at week 6.”

Revised manuscript text: “Patients also underwent a regimented pre-operative and post-operative therapy protocol. Hence, the clinical gains cannot be conclusively attributed to the surgery itself.”

L116–119

L264–266

Lines 114–116 (original).

The statistical analysis is insufficient for the study design. The use of Student’s t-tests does not appropriately account for repeated measures within subjects over time. More appropriate approaches, such as mixed-effects models or repeated-measures ANOVA, should be used. Additionally, there is no discussion of multiple comparisons or adjustment procedures.

We have rewritten the statistical methods. Within-subject pre/post comparisons now use paired Student’s t-test (continuous outcomes) and paired McNemar’s test (binary outcomes), with the Stuart-Maxwell marginal homogeneity test for the three-level scapular dyskinesis classification. The Holm-Bonferroni step-down procedure is applied across the family of paired comparisons. We considered mixed-effects modeling, but because this is a pre vs. most-recent-follow-up comparison rather than a longitudinal trajectory analysis, paired tests with Holm-Bonferroni adjustment are appropriate and we explicitly state that outcomes from the most recent visit were used. Post-hoc power for a paired Cohen’s dₘ of 0.20 is reported.

Revised manuscript text: “Paired Student’s t-test and paired McNemar’s test compared continuous and binary categorical variables of interest, respectively, while the marginal homogeneity test of Stuart-Maxwell compared the three-level scapular dyskinesis classification… Two-tailed α=0.05 was used for all tests, with the Holm-Bonferroni step-down procedure applied for multiple paired comparisons. As this study is retrospective, no a priori sample-size calculation was performed. Post hoc, with 318 paired observations, the study had >99% power to detect a paired effect of Cohen’s dₘ=0.20.”

Revised manuscript text: “Outcomes from the most recent visit were used for all analysis.”

L128–136

L123

Lines 122–128 (original).

Results are reported with uniform p-values (p < 0.01) without providing exact values, confidence intervals for changes, or standardized effect sizes. This limits interpretability and does not meet expected reporting standards.

Continuous variables are now reported throughout with 95% CIs in the text and in Tables 1 and 2. Exact p-values are reported for each outcome in Table 2 (rather than a uniform threshold) — e.g., p<0.001 for most outcomes but p=0.650 for cubital tunnel neuropathy, demonstrating that the analysis distinguishes between significant and non-significant findings. Effect-size context is provided in the discussion via the MCID benchmark (mean reduction >5 points vs. MCID of 2), and post-hoc power is given for Cohen’s dₘ.

Revised manuscript text: “Mean age was 51.0 years (49.2, 52.9)… Patients presented with a mean of 3.10 (2.96, 3.23) diagnoses.”

Revised manuscript text: “However, normalizing scapular mechanics produced substantial clinical improvement, generating a mean absolute pain reduction of >5 points, more than double the minimum clinically important difference (MCID) of 2.”

L138, L142

L170–172

Tables 1–2

Lines 122–131 (original).

The magnitude of reported effects (e.g., large reductions in pain and near elimination of headaches) is unusually large for chronic pain interventions and raises concerns about regression to the mean, placebo effects, and selection bias. These alternative explanations are not sufficiently addressed.

We have added explicit acknowledgement of placebo effect, selection bias, and the influence of concurrent rehabilitation as alternative explanations for the magnitude of effect. We have also benchmarked the effect against the published MCID for pain (2 points), and we note that the study population was preselected by HDL criteria — a feature that may contribute to the unusually large response and that we now flag as a fundamental limitation.

Revised manuscript text: “Patient selection bias and other confounders especially the placebo effect are influencing the results. Patients also underwent a regimented pre-operative and post-operative therapy protocol. Hence, the clinical gains cannot be conclusively attributed to the surgery itself.”

Revised manuscript text: “Crucially, our findings only apply to those ‘in the loop’. The terminal symptoms of the HDL can occur independently.”

L263–266

L276–277

Lines 124–125 (original).

The near-complete normalization of scapular dyskinesis suggests potential measurement bias, particularly given the lack of blinding. The manuscript does not report whether outcome assessors were blinded, which is critical in subjective clinical evaluations.

We acknowledge that formal blinded outcome assessment was not performed. To partially mitigate single-observer bias, all pre-operative exam findings were independently confirmed by a certified hand therapist (CHT). Post-operative outcomes were assessed by the treating team. We have explicitly added this as a limitation and noted that future prospective studies should include blinded independent outcome assessors with validated PROs.

Revised manuscript text: “All patients trialed at least 6 weeks of therapy before surgery and received an independent evaluation by a certified hand therapist (CHT) confirming all exam findings including coracoid tenderness, scapular protraction, and SCT+ neuropathic lesions.”

Revised manuscript text: “We did not employ standardized patient-reported outcomes (PROs), and validated instruments such as DASH or SF-36 would better capture impact on overall quality of life.”

L108–110

L274–276

Lines 128–129 (original).

The fact that 17% of patients required secondary neurolysis is clinically significant and suggests that the initial intervention may not be sufficient in a substantial subset of patients. This should be more critically discussed as a limitation rather than presented descriptively.

We have repositioned the 17% secondary neurolysis rate as a substantive limitation rather than a descriptive observation. We discuss it as evidence of the prevalence of double-crush neuropathy in this population, note that surgeons must follow explicit diagnostic criteria and counsel patients accordingly, and emphasize it in the conclusion so that prospective patients understand the staged nature of treatment.

Revised manuscript text: “17% of patients required secondary neurolysis, emphasizing the prevalence of double crush neuropathy. Surgeons should follow explicit diagnostic criteria (Figure 2), exhaust conservative options first, and closely survey patients after treatment. Breaking the loop via PM+ICN is the first step in a longitudinal and multi-disciplinary process to pain relief.”

Revised manuscript text: “Certain intractable patients may benefit substantially from PM+ICN, but should be counseled that 17% require secondary neurolysis.”

L277–281

L292–293

Lines 221–224 (original).

The discussion advances a unifying mechanistic explanation across multiple heterogeneous disorders. Given the study design, such claims are not supported and should be reframed as speculative or hypothesis-generating.

We have systematically softened mechanistic and unifying claims throughout the discussion. Conditional qualifiers (“theoretically,” “may,” “hypothesize,” “proposed model”) have been inserted at every cross-disorder claim. The closing discussion paragraph and conclusion now explicitly position HDL as a hypothesis-generating proposed model requiring further substantiation.

Revised manuscript text: “This suggests the PM pathologizing the scapula’s connections may constitute a shared anatomic mechanism that contributes to some of the intractable symptoms seen in these disorders.”

Revised manuscript text: “The HDL should be seen as a proposed model which may contribute to unexplained chronic pain spanning various disorders, but ultimately needs substantiation via larger, multi-institutional studies that include a control group, utilize PROs, and incorporate objective testing to demonstrate mechanism of action.”

L175–177

L283–286

Lines 226–234 (original).

Although limitations are acknowledged, they are presented as secondary rather than fundamental. The absence of a control group, lack of randomization, absence of validated outcome measures, and lack of blinding represent core methodological limitations that must be emphasized more strongly.

The limitations section has been substantially expanded and repositioned to emphasize that the absence of a control group, lack of randomization, absence of validated PROs, and the single-observer/non-blinded outcome assessment are fundamental — not secondary — limitations. The opening sentence of the limitations now reads “Our study suffers numerous limitations. Chiefly, this is a retrospective study limited to one practice without a control group.”

Revised manuscript text: “Our study suffers numerous limitations. Chiefly, this is a retrospective study limited to one practice without a control group. Patient selection bias and other confounders especially the placebo effect are influencing the results. Patients also underwent a regimented pre-operative and post-operative therapy protocol. Hence, the clinical gains cannot be conclusively attributed to the surgery itself… We did not employ standardized patient-reported outcomes (PROs), and validated instruments such as DASH or SF-36 would better capture impact on overall quality of life.”

L262–276

Minor Revisions

Lines 10–18 (original).

The abstract contains speculative and non-scientific language (e.g., “villainous roles”) and evolutionary explanations that are not directly supported by the data. The tone should be more neutral and aligned with scientific reporting standards.

We have removed the colloquial and dramatic phrasing (e.g., “villainous roles”) and rewritten the abstract Background in neutral, descriptive language. Evolutionary framing has been moved out of the abstract entirely; it remains only in the Discussion as a clearly demarcated hypothesis-generating section.

Revised manuscript text: “Many patients suffer from chronic pain of the shoulder, neck, upper back, and/or arm… The pectoralis minor (PM) is the only muscle of the scapula controlled by the lower trunk of the brachial plexus. In the Human Disharmony Loop (HDL), this neurologic asymmetry produces persistent protraction of the scapula. Protraction deforms the scapula’s connections, generating headaches and neck stiffness, upper back tightness, shoulder weakness, and hand numbness. We hypothesize patients with the above who meet HDL diagnostic criteria will benefit from PM tenotomy with brachial plexus neurolysis (PM+ICN).”

L17–26

Lines 40–52 (original).

The introduction emphasizes the intractability of chronic pain conditions without sufficiently acknowledging variability in treatment outcomes or existing effective interventions. A more balanced presentation of the literature is needed.

We have softened the introductory language from absolute (“ineffectual treatments”) to qualified (“sometimes ineffectual treatments”) and replaced absolute disability framing with descriptive language (“remains mysterious and challenging”). We acknowledge that variability in outcomes exists and that some patients respond well to existing interventions.

Revised manuscript text: “Despite its ubiquity, chronic pain afflicting the upper limb girdle – neck, upper back, shoulder, and arm – remains mysterious and challenging… They also share contentious diagnostic criteria, convoluted pathophysiology, and sometimes ineffectual treatments.”

L49–57

Lines 63–74 (original).

The HDL model is presented as a coherent explanatory framework without clearly distinguishing between hypothesis and established evidence. This distinction should be clarified.

We have inserted explicit “theoretically” and “we hypothesize” qualifiers when introducing HDL and its anatomic links to chronic pain disorders, and we have clearly stated that the connection between HDL and the various chronic pain syndromes is a theoretical proposal rather than established evidence.

Revised manuscript text: “Theoretically, the HDL anatomically connects to the above chronic pain disorders (Figure 3). Hence, we hypothesize the PM pathologizing scapular mechanics may contribute to refractory symptoms in these conditions.”

L77–80

Lines 100–101 (original).

The manuscript refers to a self-reported pain questionnaire but does not clearly specify whether a standardized instrument (e.g., VAS or NRS) was used. This should be clarified for reproducibility.

We have clarified that the standardized 0–10 Visual Analogue Scale (VAS) was used for pain assessment, both in the abstract and in the methods.

Revised manuscript text: “At each visit, patients prospectively completed a self-reported Visual Analogue Scale (VAS) pain questionnaire, which active shoulder abduction range of motion (ROM) values were measured.”

Revised manuscript text: “Outcomes included self-reported Visual Analogue Score pain scores, active shoulder abduction range of motion (ROM), prevalence of occipital headaches.”

L111–113

L30–31

Lines 109–111 (original).

The categorization of treatment response (poor, important, substantial) appears arbitrary and lacks statistical justification. The authors should provide references or consider removing this classification.

This categorical treatment-response classification (substantial / important / poor) was based on IMMPACT guidelines recommended by pain societies. However, we agree with this reviewer and have decided to remove it entirely. We now report only continuous outcomes for self-reported pain.

Table 1 (Lines 119–132 original).

While descriptive statistics are presented clearly, there is no exploration of variability across patient subgroups, which limits interpretability.

Table 1 retains the descriptive baseline characteristics. We have added Figure 5 (forest plot of pain reduction with 95% CIs by diagnosis) and a formal cross-group heterogeneity test (Cochran’s Q, I²) to characterize variability across subgroups. Prior-surgery subcategories are also reported in Table 1.

L154–156, L161

Table 1

Figure 5 (Page 6 original).

The figure presents strong effects but lacks detailed statistical annotation, including confidence intervals and clarification of error bars.

The revised Figure 5 (forest plot of pain reduction by diagnosis) explicitly displays mean pain reduction with 95% CI error bars per diagnosis and reports the cross-group Cochran’s Q test and I² statistic directly on the figure. The legend has been clarified to describe the statistical content.

Revised manuscript text: “Figure 5. Pain Reductions by Chronic Pain Diagnosis. The absolute pain reduction among the heterogenous chronic pain diagnoses was statistically similar, consistent with a shared underlying mechanism.”

L161, L163–164

Figure 6 (Page 7 original).

The distribution of treatment response is visually informative, but the statistical interpretation of the density plot is not explained, limiting its utility.

We have removed this Figure entirely.

Lines 232–233 (original).

The absence of validated patient-reported outcome measures is acknowledged but should be emphasized as a major limitation rather than a minor point.

We have repositioned the absence of validated PROs as a major limitation within the expanded limitations section. We explicitly call out DASH and SF-36 as instruments that future studies should employ, and we cite the relevant PRO methodology literature.

Revised manuscript text: “We did not employ standardized patient-reported outcomes (PROs), and validated instruments such as DASH or SF-36 would better capture impact on overall quality of life.”

Revised manuscript text: “…ultimately needs substantiation via larger, multi-institutional studies that include a control group, utilize PROs, and incorporate objective testing to demonstrate mechanism of action.”

L274–276

L283–286

References (Lines 267–361 original).

While generally appropriate, some references are older or narrative in nature. Where possible, more recent high-quality systematic reviews should be incorporated.

We have updated the reference list with recent high-quality systematic reviews where available. New additions include Faszholz & Cheng (2024) systematic review on the scratch-collapse test (ref 46), Gabriel et al. (2025) on confirmatory testing in carpal tunnel syndrome (ref 47), Campbell et al. (2022) systematic review on PROs in clinical practice (ref 49), Zini & Banfi (2021) on bias in PROMs (ref 50), Karjalainen et al. (2019) Cochrane review on subacromial decompression (ref 6), and Povlsen et al. (2014) Cochrane review on TOS (ref 4). Older narrative references have been retained only where they remain the seminal source for the concept cited.

Refs 4, 6, 46–50

Closing Remarks.

The manuscript addresses an interesting and potentially important clinical concept but currently overinterprets findings derived from a methodologically limited design. Substantial revisions are required to improve statistical rigor, clarify methodology, and align conclusions with the level of evidence.

We are grateful to Reviewer 2 for the rigorous and constructive critique. We believe the substantially revised manuscript — with explicit retrospective observational framing, hypothesis-generating language throughout, expanded statistical methods (paired tests, 95% CIs, Holm-Bonferroni adjustment, Cochran’s Q heterogeneity, post-hoc power), per-diagnosis subgroup analysis, complications reporting, and an expanded limitations section — now appropriately aligns its conclusions with the level of evidence the study can support. We remain enthusiastic about the HDL framework as a hypothesis-generating contribution and welcome further suggestions.

Throughout