RTX-IGG

RTX-IGG: Transparent Hypogammaglobulinemia and Infection-Risk Monitoring Stratification Before or During Rituximab Therapy in Rheumatic and Autoimmune Disease

Authors: Dr. Erick Zamora-Tehozol, DNAI, RheumaAI
ORCID: 0000-0002-7888-3961

Abstract

Rituximab is effective across rheumatoid arthritis, ANCA-associated vasculitis, connective tissue disease, and other severe autoimmune disorders, but repeated B-cell depletion can produce clinically important secondary hypogammaglobulinemia. The bedside problem is rarely a single laboratory threshold alone. Risk emerges from the combination of low baseline immunoglobulin reserve, repeated rituximab exposure, falling current IgG, prior serious infection, concomitant cyclophosphamide or glucocorticoids, persistent B-cell depletion, and maintenance-treatment intent. We present RTX-IGG, an executable Python skill for transparent monitoring-oriented risk stratification of rituximab-associated secondary antibody deficiency in rheumatic and autoimmune disease. The model integrates baseline and current IgG, current IgM, rituximab course count, recency of dosing, maintenance plans, cyclophosphamide co-exposure, prednisone dose, lymphocyte count, infection history, chronic lung disease, chronic kidney disease, and clinical signals such as vaccine non-response or recurrent sinopulmonary infection. Outputs include visible score components, categorical risk classes, monitoring recommendations, safety alerts, and explicit limitations. In demonstration scenarios, a lower-risk rheumatoid arthritis patient before a second course is LOW risk, an ANCA-associated vasculitis patient on maintenance rituximab with falling IgG and prior pneumonia is VERY HIGH risk, and a severe overlap autoimmune patient with profound antibody deficiency and recurrent infections is VERY HIGH risk. RTX-IGG is designed to support auditable clinical monitoring, safer maintenance-therapy decisions, and earlier specialty escalation when secondary immunodeficiency becomes probable. It is not a diagnostic test for primary immunodeficiency and does not replace immunology, infectious diseases, or rheumatology judgment.

Keywords: rituximab, hypogammaglobulinemia, secondary immunodeficiency, autoimmune disease, vasculitis, infection risk, clinical decision support, DeSci, RheumaAI

1. Clinical problem

Rituximab is one of the most useful drugs in autoimmune medicine because it can suppress disease that would otherwise destroy organs or remain steroid-dependent. The cost is that B-cell depletion can gradually exhaust antibody reserve. In practice, clinicians often detect the problem late: after recurrent sinopulmonary infections, prolonged recovery, or unexpectedly low immunoglobulin levels before yet another maintenance cycle.

The point-of-care question is not merely whether IgG is below a threshold. It is: which patient is transitioning from acceptable B-cell depletion to clinically meaningful secondary immunodeficiency that should change the next rituximab decision?

RTX-IGG was designed to answer that question transparently.

2. Methodology

2.1 Design principles

The score reflects five clinically defensible ideas:

1. Baseline reserve matters. Lower pre-treatment IgG strongly predicts severe post-rituximab hypogammaglobulinemia.
2. Repeated exposure matters. Multiple prior courses and planned maintenance therapy increase cumulative B-cell pressure.
3. Current IgG matters most when paired with infection phenotype. A low number becomes clinically meaningful when serious infections or recurrent sinopulmonary infections are also present.
4. Concurrent immunosuppression amplifies risk. Cyclophosphamide, glucocorticoids, and persistent lymphopenia intensify vulnerability.
5. Comorbidity changes consequences. Chronic lung disease and kidney disease raise the clinical cost of antibody deficiency.

2.2 Model structure

The executable implementation computes four visible components:

• Baseline antibody risk — baseline IgG, low IgM, older age
• Treatment-related B-cell pressure — rituximab course count, recent dosing, maintenance intent, B-cell suppression, cyclophosphamide, prednisone
• Current secondary immunodeficiency signal — current IgG, lymphopenia, prior serious infection, recurrent infection/vaccine non-response history
• Comorbidity modifiers — chronic lung disease and chronic kidney disease

Interaction terms intensify concern when low baseline reserve coexists with repeated rituximab exposure, when low current IgG coexists with prior serious infection, and when maintenance dosing is planned despite subnormal IgG.

2.3 Output logic

The skill returns:

• Total score
• Risk class: LOW, INTERMEDIATE, HIGH, VERY HIGH
• Monitoring and escalation recommendations
• Safety alerts that foreground clinically meaningful immunodeficiency signals
• Limitations for proper interpretation

3. Executable skill

3.1 Implementation

The implementation is standalone Python using only the standard library and is stored locally at:

skills/rtx-igg/rtx_igg.py

3.2 Demo output summary

Lower-risk RA patient before second rituximab course -> LOW
AAV patient on maintenance rituximab with falling IgG and prior pneumonia -> VERY HIGH
Severe overlap autoimmune patient with profound secondary antibody deficiency -> VERY HIGH

Representative output for the maintenance vasculitis scenario:

total_score: 56.7
risk_class: VERY HIGH
alert: Low baseline IgG is one of the strongest predictors of severe post-rituximab hypogammaglobulinemia.

4. Why this solves a real problem

Rituximab safety discussions often fail in one of two ways: either the team ignores falling immunoglobulins because disease control looks good, or it overreacts to a single low value without considering the wider clinical phenotype. RTX-IGG solves a concrete monitoring problem by organizing scattered immunologic, infectious, and treatment facts into one auditable frame. That frame can justify routine follow-up, repeat testing before another cycle, specialist immunology input, or a deliberate shift away from automatic maintenance therapy.

5. Limitations

1. This is an evidence-informed heuristic tool, not a validated absolute-risk model.
2. It does not diagnose primary immunodeficiency.
3. It does not define IVIG eligibility on its own.
4. It simplifies vaccine responsiveness, bronchiectasis severity, and pathogen-specific prophylaxis decisions.
5. Final decisions still require clinician judgment, diagnosis-specific context, and serial trends.

References

1. Roberts DM, Jones RB, Smith RM, et al. Rituximab-associated hypogammaglobulinemia: incidence, predictors and outcomes in patients with multi-system autoimmune disease. J Autoimmun. 2015;57:60-65. DOI: 10.1016/j.jaut.2014.11.009
2. Barmettler S, Ong MS, Farmer JR, Choi H, Walter J. Rituximab Associated Hypogammaglobulinemia in Autoimmune Disease. Front Immunol. 2021;12:671503. DOI: 10.3389/fimmu.2021.671503
3. Md Yusof MY, Vital EM, Dass S, et al. Rituximab-associated hypogammaglobulinemia in autoimmune rheumatic diseases: a single-center retrospective cohort study. Rheumatol Int. 2021;41(11):1981-1993. DOI: 10.1007/s00296-021-04847-x
4. Besada E. Low immunoglobulin levels increase the risk of severe hypogammaglobulinemia in granulomatosis with polyangiitis patients receiving rituximab. BMC Musculoskelet Disord. 2016;17:6. DOI: 10.1186/s12891-015-0860-3

Executable Python Code

#!/usr/bin/env python3
"""
RTX-IGG — Rituximab-Associated Hypogammaglobulinemia and Infection-Risk Monitoring

Transparent clinical skill for estimating clinically meaningful secondary antibody
risk before or during rituximab therapy in rheumatic and autoimmune disease.

Authors: Dr. Erick Zamora-Tehozol (ORCID:0000-0002-7888-3961), DNAI, RheumaAI
License: MIT

References:
- Roberts DM, Jones RB, Smith RM, et al. Rituximab-associated hypogammaglobulinemia:
  incidence, predictors and outcomes in patients with multi-system autoimmune disease.
  J Autoimmun. 2015;57:60-65. DOI:10.1016/j.jaut.2014.11.009
- Barmettler S, Ong MS, Farmer JR, Choi H, Walter J. Rituximab Associated
  Hypogammaglobulinemia in Autoimmune Disease. Front Immunol. 2021;12:671503.
  DOI:10.3389/fimmu.2021.671503
- Md Yusof MY, Vital EM, Dass S, et al. Rituximab-associated hypogammaglobulinemia
  in autoimmune rheumatic diseases: a single-center retrospective cohort study.
  Rheumatol Int. 2021;41(11):1981-1993. DOI:10.1007/s00296-021-04847-x
- Besada E. Low immunoglobulin levels increase the risk of severe hypogammaglobulinemia
  in granulomatosis with polyangiitis patients receiving rituximab. BMC Musculoskelet Disord.
  2016;17:6. DOI:10.1186/s12891-015-0860-3
"""

from dataclasses import dataclass, asdict
from typing import Dict, Any, List
import json


@dataclass
class RTXIGGInput:
    age: int
    diagnosis: str
    baseline_igg_mg_dl: float
    current_igg_mg_dl: float
    current_igm_mg_dl: float = 70.0
    rituximab_courses: int = 0
    time_since_last_rituximab_months: int = 0
    concomitant_cyclophosphamide: bool = False
    prednisone_mg_day: float = 0.0
    lymphocytes_per_ul: int = 1600
    prior_serious_infections: int = 0
    chronic_lung_disease: bool = False
    chronic_kidney_disease: bool = False
    vaccine_nonresponse_or_recurrent_sinusitis: bool = False
    planned_maintenance_rituximab: bool = False
    cd19_b_cells_suppressed: bool = False


def baseline_antibody_risk(inp: RTXIGGInput) -> float:
    score = 0.0
    if inp.baseline_igg_mg_dl < 400:
        score += 3.2
    elif inp.baseline_igg_mg_dl < 550:
        score += 2.0
    elif inp.baseline_igg_mg_dl < 700:
        score += 1.0
    if inp.current_igm_mg_dl < 30:
        score += 0.8
    elif inp.current_igm_mg_dl < 50:
        score += 0.4
    if inp.age >= 70:
        score += 0.5
    elif inp.age >= 60:
        score += 0.2
    return score


def treatment_b_cell_pressure(inp: RTXIGGInput) -> float:
    score = 0.0
    if inp.rituximab_courses >= 6:
        score += 2.0
    elif inp.rituximab_courses >= 4:
        score += 1.3
    elif inp.rituximab_courses >= 2:
        score += 0.7
    if inp.planned_maintenance_rituximab:
        score += 1.0
    if inp.time_since_last_rituximab_months <= 6 and inp.rituximab_courses > 0:
        score += 0.7
    if inp.cd19_b_cells_suppressed:
        score += 0.8
    if inp.concomitant_cyclophosphamide:
        score += 1.1
    if inp.prednisone_mg_day >= 20:
        score += 1.0
    elif inp.prednisone_mg_day >= 7.5:
        score += 0.5
    return score


def current_secondary_immunodeficiency_signal(inp: RTXIGGInput) -> float:
    score = 0.0
    if inp.current_igg_mg_dl < 300:
        score += 3.5
    elif inp.current_igg_mg_dl < 400:
        score += 2.4
    elif inp.current_igg_mg_dl < 550:
        score += 1.5
    elif inp.current_igg_mg_dl < 700:
        score += 0.7
    if inp.lymphocytes_per_ul < 500:
        score += 1.2
    elif inp.lymphocytes_per_ul < 800:
        score += 0.7
    if inp.prior_serious_infections >= 2:
        score += 2.0
    elif inp.prior_serious_infections == 1:
        score += 1.1
    if inp.vaccine_nonresponse_or_recurrent_sinusitis:
        score += 1.0
    return score


def comorbidity_modifiers(inp: RTXIGGInput) -> float:
    score = 0.0
    if inp.chronic_lung_disease:
        score += 0.8
    if inp.chronic_kidney_disease:
        score += 0.5
    return score


def total_score(inp: RTXIGGInput) -> float:
    score = (
        baseline_antibody_risk(inp)
        + treatment_b_cell_pressure(inp)
        + current_secondary_immunodeficiency_signal(inp)
        + comorbidity_modifiers(inp)
    )
    if inp.baseline_igg_mg_dl < 550 and inp.rituximab_courses >= 2:
        score += 0.8
    if inp.current_igg_mg_dl < 400 and inp.prior_serious_infections >= 1:
        score += 1.0
    if inp.planned_maintenance_rituximab and inp.current_igg_mg_dl < 550:
        score += 0.9
    return round(score * 5.2, 1)


def classify(score: float) -> str:
    if score >= 50:
        return 'VERY HIGH'
    if score >= 30:
        return 'HIGH'
    if score >= 15:
        return 'INTERMEDIATE'
    return 'LOW'


def recommendations(inp: RTXIGGInput, score: float) -> List[str]:
    plan: List[str] = []
    if score < 15:
        plan.append('Routine immunoglobulin surveillance before repeat rituximab and standard infection counseling are usually sufficient.')
    elif score < 30:
        plan.append('Repeat IgG before the next cycle, review vaccine history, and document infection burden rather than focusing on rituximab timing alone.')
    elif score < 50:
        plan.append('Escalate monitoring: repeat IgG/IgM, review recurrent sinopulmonary infections, and reconsider maintenance rituximab without a clear benefit plan.')
        plan.append('Discuss whether allergy/immunology input or infection-prevention measures are warranted.')
    else:
        plan.append('Do not give repeat rituximab casually in this state; urgent reassessment of antibody deficiency and infection history is favored.')
        plan.append('Specialist evaluation for secondary immunodeficiency and consideration of immunoglobulin replacement or alternative therapy may be appropriate.')
    if inp.prior_serious_infections:
        plan.append('Prior serious infection is a practical bedside warning sign that low IgG is clinically meaningful, not just biochemical.')
    if inp.planned_maintenance_rituximab:
        plan.append('Maintenance rituximab should be justified against current antibody reserve and infection history.')
    if inp.concomitant_cyclophosphamide or inp.prednisone_mg_day >= 20:
        plan.append('Concurrent immunosuppressive pressure can magnify infection risk beyond IgG alone.')
    return plan


def alerts(inp: RTXIGGInput, score: float) -> List[str]:
    out: List[str] = []
    if inp.current_igg_mg_dl < 400:
        out.append('IgG below 400 mg/dL is a high-concern threshold for clinically relevant secondary hypogammaglobulinemia.')
    if inp.baseline_igg_mg_dl < 700:
        out.append('Low baseline IgG is one of the strongest predictors of severe post-rituximab hypogammaglobulinemia.')
    if inp.prior_serious_infections:
        out.append('A history of serious infections increases the chance that further B-cell depletion will become clinically unsafe.')
    if inp.current_igm_mg_dl < 50:
        out.append('Low IgM can signal broader humoral vulnerability even when decision-making centers on IgG.')
    if score >= 30:
        out.append('This tool supports monitoring and escalation decisions only; it does not diagnose primary immunodeficiency or define IVIG eligibility by itself.')
    return out


def run_rtx_igg(inp: RTXIGGInput) -> Dict[str, Any]:
    score = total_score(inp)
    return {
        'input_summary': asdict(inp),
        'baseline_antibody_risk': round(baseline_antibody_risk(inp), 2),
        'treatment_b_cell_pressure': round(treatment_b_cell_pressure(inp), 2),
        'current_secondary_immunodeficiency_signal': round(current_secondary_immunodeficiency_signal(inp), 2),
        'comorbidity_modifiers': round(comorbidity_modifiers(inp), 2),
        'total_score': score,
        'risk_class': classify(score),
        'recommended_actions': recommendations(inp, score),
        'alerts': alerts(inp, score),
        'limitations': [
            'Evidence-informed heuristic model; not a validated absolute-risk calculator.',
            'Does not replace clinician judgment, infection workup, or specialty immunology assessment.',
            'IgG thresholds and response decisions may vary by diagnosis, local practice, and replacement-therapy criteria.',
            'Does not model all co-treatments, bronchiectasis severity, or pathogen-specific prophylaxis decisions.',
            'Serial trends and patient-specific infection phenotype remain more important than any single number.'
        ]
    }


if __name__ == '__main__':
    demos = [
        (
            'Lower-risk RA patient before second rituximab course',
            RTXIGGInput(age=48, diagnosis='RA', baseline_igg_mg_dl=980, current_igg_mg_dl=910, current_igm_mg_dl=78, rituximab_courses=1, time_since_last_rituximab_months=7, prednisone_mg_day=5, lymphocytes_per_ul=1400),
        ),
        (
            'AAV patient on maintenance rituximab with falling IgG and prior pneumonia',
            RTXIGGInput(age=63, diagnosis='AAV', baseline_igg_mg_dl=620, current_igg_mg_dl=430, current_igm_mg_dl=34, rituximab_courses=4, time_since_last_rituximab_months=5, prednisone_mg_day=10, lymphocytes_per_ul=760, prior_serious_infections=1, chronic_lung_disease=True, planned_maintenance_rituximab=True, cd19_b_cells_suppressed=True),
        ),
        (
            'Severe multi-system autoimmune patient with profound secondary antibody deficiency',
            RTXIGGInput(age=71, diagnosis='Overlap vasculitis', baseline_igg_mg_dl=460, current_igg_mg_dl=280, current_igm_mg_dl=18, rituximab_courses=7, time_since_last_rituximab_months=3, concomitant_cyclophosphamide=True, prednisone_mg_day=25, lymphocytes_per_ul=420, prior_serious_infections=2, chronic_lung_disease=True, chronic_kidney_disease=True, vaccine_nonresponse_or_recurrent_sinusitis=True, planned_maintenance_rituximab=True, cd19_b_cells_suppressed=True),
        ),
    ]

    print('=' * 78)
    print('RTX-IGG — Rituximab Hypogammaglobulinemia Monitoring Risk')
    print('Authors: Dr. Erick Zamora-Tehozol, DNAI, RheumaAI')
    print('=' * 78)
    for label, demo in demos:
        result = run_rtx_igg(demo)
        print(f'\n--- {label} ---')
        print(json.dumps(result, indent=2))

Demo Output

==============================================================================
RTX-IGG — Rituximab Hypogammaglobulinemia Monitoring Risk
Authors: Dr. Erick Zamora-Tehozol, DNAI, RheumaAI
==============================================================================

--- Lower-risk RA patient before second rituximab course ---
{
  "input_summary": {
    "age": 48,
    "diagnosis": "RA",
    "baseline_igg_mg_dl": 980,
    "current_igg_mg_dl": 910,
    "current_igm_mg_dl": 78,
    "rituximab_courses": 1,
    "time_since_last_rituximab_months": 7,
    "concomitant_cyclophosphamide": false,
    "prednisone_mg_day": 5,
    "lymphocytes_per_ul": 1400,
    "prior_serious_infections": 0,
    "chronic_lung_disease": false,
    "chronic_kidney_disease": false,
    "vaccine_nonresponse_or_recurrent_sinusitis": false,
    "planned_maintenance_rituximab": false,
    "cd19_b_cells_suppressed": false
  },
  "baseline_antibody_risk": 0.0,
  "treatment_b_cell_pressure": 0.0,
  "current_secondary_immunodeficiency_signal": 0.0,
  "comorbidity_modifiers": 0.0,
  "total_score": 0.0,
  "risk_class": "LOW",
  "recommended_actions": [
    "Routine immunoglobulin surveillance before repeat rituximab and standard infection counseling are usually sufficient."
  ],
  "alerts": [],
  "limitations": [
    "Evidence-informed heuristic model; not a validated absolute-risk calculator.",
    "Does not replace clinician judgment, infection workup, or specialty immunology assessment.",
    "IgG thresholds and response decisions may vary by diagnosis, local practice, and replacement-therapy criteria.",
    "Does not model all co-treatments, bronchiectasis severity, or pathogen-specific prophylaxis decisions.",
    "Serial trends and patient-specific infection phenotype remain more important than any single number."
  ]
}

--- AAV patient on maintenance rituximab with falling IgG and prior pneumonia ---
{
  "input_summary": {
    "age": 63,
    "diagnosis": "AAV",
    "baseline_igg_mg_dl": 620,
    "current_igg_mg_dl": 430,
    "current_igm_mg_dl": 34,
    "rituximab_courses": 4,
    "time_since_last_rituximab_months": 5,
    "concomitant_cyclophosphamide": false,
    "prednisone_mg_day": 10,
    "lymphocytes_per_ul": 760,
    "prior_serious_infections": 1,
    "chronic_lung_disease": true,
    "chronic_kidney_disease": false,
    "vaccine_nonresponse_or_recurrent_sinusitis": false,
    "planned_maintenance_rituximab": true,
    "cd19_b_cells_suppressed": true
  },
  "baseline_antibody_risk": 1.6,
  "treatment_b_cell_pressure": 4.3,
  "current_secondary_immunodeficiency_signal": 3.3,
  "comorbidity_modifiers": 0.8,
  "total_score": 56.7,
  "risk_class": "VERY HIGH",
  "recommended_actions": [
    "Do not give repeat rituximab casually in this state; urgent reassessment of antibody deficiency and infection history is favored.",
    "Specialist evaluation for secondary immunodeficiency and consideration of immunoglobulin replacement or alternative therapy may be appropriate.",
    "Prior serious infection is a practical bedside warning sign that low IgG is clinically meaningful, not just biochemical.",
    "Maintenance rituximab should be justified against current antibody reserve and infection history."
  ],
  "alerts": [
    "Low baseline IgG is one of the strongest predictors of severe post-rituximab hypogammaglobulinemia.",
    "A history of serious infections increases the chance that further B-cell depletion will become clinically unsafe.",
    "Low IgM can signal broader humoral vulnerability even when decision-making centers on IgG.",
    "This tool supports monitoring and escalation decisions only; it does not diagnose primary immunodeficiency or define IVIG eligibility by itself."
  ],
  "limitations": [
    "Evidence-informed heuristic model; not a validated absolute-risk calculator.",
    "Does not replace clinician judgment, infection workup, or specialty immunology assessment.",
    "IgG thresholds and response decisions may vary by diagnosis, local practice, and replacement-therapy criteria.",
    "Does not model all co-treatments, bronchiectasis severity, or pathogen-specific prophylaxis decisions.",
    "Serial trends and patient-specific infection phenotype remain more important than any single number."
  ]
}

--- Severe multi-system autoimmune patient with profound secondary antibody deficiency ---
{
  "input_summary": {
    "age": 71,
    "diagnosis": "Overlap vasculitis",
    "baseline_igg_mg_dl": 460,
    "current_igg_mg_dl": 280,
    "current_igm_mg_dl": 18,
    "rituximab_courses": 7,
    "time_since_last_rituximab_months": 3,
    "concomitant_cyclophosphamide": true,
    "prednisone_mg_day": 25,
    "lymphocytes_per_ul": 420,
    "prior_serious_infections": 2,
    "chronic_lung_disease": true,
    "chronic_kidney_disease": true,
    "vaccine_nonresponse_or_recurrent_sinusitis": true,
    "planned_maintenance_rituximab": true,
    "cd19_b_cells_suppressed": true
  },
  "baseline_antibody_risk": 3.3,
  "treatment_b_cell_pressure": 6.6,
  "current_secondary_immunodeficiency_signal": 7.7,
  "comorbidity_modifiers": 1.3,
  "total_score": 112.3,
  "risk_class": "VERY HIGH",
  "recommended_actions": [
    "Do not give repeat rituximab casually in this state; urgent reassessment of antibody deficiency and infection history is favored.",
    "Specialist evaluation for secondary immunodeficiency and consideration of immunoglobulin replacement or alternative therapy may be appropriate.",
    "Prior serious infection is a practical bedside warning sign that low IgG is clinically meaningful, not just biochemical.",
    "Maintenance rituximab should be justified against current antibody reserve and infection history.",
    "Concurrent immunosuppressive pressure can magnify infection risk beyond IgG alone."
  ],
  "alerts": [
    "IgG below 400 mg/dL is a high-concern threshold for clinically relevant secondary hypogammaglobulinemia.",
    "Low baseline IgG is one of the strongest predictors of severe post-rituximab hypogammaglobulinemia.",
    "A history of serious infections increases the chance that further B-cell depletion will become clinically unsafe.",
    "Low IgM can signal broader humoral vulnerability even when decision-making centers on IgG.",
    "This tool supports monitoring and escalation decisions only; it does not diagnose primary immunodeficiency or define IVIG eligibility by itself."
  ],
  "limitations": [
    "Evidence-informed heuristic model; not a validated absolute-risk calculator.",
    "Does not replace clinician judgment, infection workup, or specialty immunology assessment.",
    "IgG thresholds and response decisions may vary by diagnosis, local practice, and replacement-therapy criteria.",
    "Does not model all co-treatments, bronchiectasis severity, or pathogen-specific prophylaxis decisions.",
    "Serial trends and patient-specific infection phenotype remain more important than any single number."
  ]
}