import json, subprocess, base64, sys
from io import BytesIO
from pptx import Presentation
from pptx.util import Inches, Pt, Emu
from pptx.dml.color import RGBColor
from pptx.enum.text import PP_ALIGN, MSO_ANCHOR
from pptx.oxml.ns import qn
from lxml import etree

# ── Colours ──────────────────────────────────────────────────────────────
DARK_BLUE   = RGBColor(0x1A, 0x37, 0x6C)   # deep navy
MID_BLUE    = RGBColor(0x2E, 0x6D, 0xB4)   # accent blue
LIGHT_BLUE  = RGBColor(0xD6, 0xE8, 0xF7)   # slide background tint
RED_ACCENT  = RGBColor(0xC0, 0x39, 0x2B)   # highlight
WHITE       = RGBColor(0xFF, 0xFF, 0xFF)
DARK_GREY   = RGBColor(0x2C, 0x2C, 0x2C)
MID_GREY    = RGBColor(0x55, 0x55, 0x55)
LIGHT_GREY  = RGBColor(0xF2, 0xF4, 0xF7)

# ── Helper utilities ──────────────────────────────────────────────────────
def add_rect(slide, l, t, w, h, fill_rgb=None, line_rgb=None, line_width_pt=0):
    shape = slide.shapes.add_shape(1, Inches(l), Inches(t), Inches(w), Inches(h))
    shape.line.fill.background()
    if fill_rgb:
        shape.fill.solid()
        shape.fill.fore_color.rgb = fill_rgb
    else:
        shape.fill.background()
    if line_rgb and line_width_pt:
        shape.line.color.rgb = line_rgb
        shape.line.width = Pt(line_width_pt)
    else:
        shape.line.fill.background()
    return shape

def add_tb(slide, l, t, w, h, text, size, bold=False, color=DARK_GREY,
           align=PP_ALIGN.LEFT, wrap=True, italic=False, font="Calibri"):
    tb = slide.shapes.add_textbox(Inches(l), Inches(t), Inches(w), Inches(h))
    tf = tb.text_frame
    tf.word_wrap = wrap
    tf.margin_left = 0; tf.margin_right = 0
    tf.margin_top = 0;  tf.margin_bottom = 0
    p = tf.paragraphs[0]
    p.alignment = align
    run = p.add_run()
    run.text = text
    run.font.name = font
    run.font.size = Pt(size)
    run.font.bold = bold
    run.font.italic = italic
    run.font.color.rgb = color
    return tf

def add_bullet_tf(slide, l, t, w, h, items, size=13, color=DARK_GREY,
                  bold_first=False, font="Calibri", line_space=1.15):
    tb = slide.shapes.add_textbox(Inches(l), Inches(t), Inches(w), Inches(h))
    tf = tb.text_frame
    tf.word_wrap = True
    tf.margin_left = 0; tf.margin_right = 0
    tf.margin_top = 0;  tf.margin_bottom = 0
    from pptx.oxml.ns import qn
    from lxml import etree
    for i, item in enumerate(items):
        if i == 0:
            p = tf.paragraphs[0]
        else:
            p = tf.add_paragraph()
        # line spacing
        pPr = p._pPr
        if pPr is None:
            pPr = p._p.get_or_add_pPr()
        lnSpc = etree.SubElement(pPr, qn('a:lnSpc'))
        spcPct = etree.SubElement(lnSpc, qn('a:spcPct'))
        spcPct.set('val', str(int(line_space * 100000)))
        spcBef = etree.SubElement(pPr, qn('a:spcBef'))
        spcPts = etree.SubElement(spcBef, qn('a:spcPts'))
        spcPts.set('val', '800')

        run = p.add_run()
        if isinstance(item, tuple):
            bullet, text = item
        else:
            bullet, text = "•  ", item
        run.text = bullet + text
        run.font.name = font
        run.font.size = Pt(size)
        run.font.color.rgb = color
        run.font.bold = (i == 0 and bold_first)
    return tf

def fetch_images(urls):
    result = json.loads(subprocess.check_output(
        ["python", "/tmp/skills/shared/scripts/fetch_images.py"] + urls
    ))
    return result

def add_image_from_b64(slide, b64_str, l, t, w, h=None):
    raw = base64.b64decode(b64_str.split(',')[1] if ',' in b64_str else b64_str)
    buf = BytesIO(raw)
    if h:
        slide.shapes.add_picture(buf, Inches(l), Inches(t), Inches(w), Inches(h))
    else:
        slide.shapes.add_picture(buf, Inches(l), Inches(t), Inches(w))

# ── Fetch all images ──────────────────────────────────────────────────────
IMAGE_URLS = [
    "https://cdn.orris.care/cdss_images/393983627f910dbe2890e4171c6754c66b496a5b5bb49f34427b6303fced4c9a.png",   # 0 X-ray Codman triangle
    "https://cdn.orris.care/cdss_images/67d666f1c3d7c28aaef14dc5976c9abb955d31bb18ec7afbac8b996a2f8e8fc6.png",   # 1 gross specimen
    "https://cdn.orris.care/cdss_images/825ac63da1176854cfe0feab96138b8c8893240b7088f04166856fa60ded9702.png",   # 2 histology
    "https://cdn.orris.care/cdss_images/6f23f9305fb931f4bfc24c17ee2e15f3095932325f0d0c8655f7628ab82b2ad5.png",   # 3 lytic/sclerotic X-rays
]
print("Fetching images...")
imgs = fetch_images(IMAGE_URLS)
print("Images fetched:", [bool(i["base64"]) for i in imgs])

# ── Presentation setup ────────────────────────────────────────────────────
prs = Presentation()
prs.slide_width  = Inches(13.333)
prs.slide_height = Inches(7.5)
blank = prs.slide_layouts[6]

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 1 – Title Slide
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, DARK_BLUE)            # full bg
add_rect(s, 0, 5.5, 13.333, 2.0, MID_BLUE)            # lower bar
add_rect(s, 0.55, 1.6, 0.07, 3.5, RED_ACCENT)         # vertical accent line

add_tb(s, 0.9, 1.4, 11.5, 1.1, "OSTEOSARCOMA", 54, bold=True,
       color=WHITE, align=PP_ALIGN.LEFT, font="Calibri")
add_tb(s, 0.9, 2.55, 11, 0.7, "Primary Malignant Bone Tumour — A Seminar", 22,
       color=RGBColor(0xAD, 0xD4, 0xF0), align=PP_ALIGN.LEFT, italic=True)

add_tb(s, 0.9, 5.7, 9, 0.5, "Department of Orthopaedics / Pathology", 15,
       color=WHITE, align=PP_ALIGN.LEFT)
add_tb(s, 0.9, 6.15, 9, 0.4, "Undergraduate MBBS Seminar  |  July 2026", 13,
       color=RGBColor(0xAD, 0xD4, 0xF0), align=PP_ALIGN.LEFT)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 2 – Overview / Contents
# ═══════════════════════════════════════════════════════════════════════════
def slide_header(slide, title, subtitle=""):
    add_rect(slide, 0, 0, 13.333, 1.15, DARK_BLUE)
    add_rect(slide, 0, 1.15, 13.333, 0.06, MID_BLUE)
    add_tb(slide, 0.45, 0.17, 12, 0.75, title, 28, bold=True,
           color=WHITE, align=PP_ALIGN.LEFT)
    if subtitle:
        add_tb(slide, 0.45, 0.82, 12, 0.38, subtitle, 13,
               color=RGBColor(0xAD, 0xD4, 0xF0), align=PP_ALIGN.LEFT, italic=True)

s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Outline of Seminar")

topics = [
    ("01", "Definition & Introduction"),
    ("02", "Epidemiology & Risk Factors"),
    ("03", "Classification (WHO 2013)"),
    ("04", "Pathogenesis & Molecular Genetics"),
    ("05", "Clinical Features"),
    ("06", "Radiological Features"),
    ("07", "Pathology (Gross & Histology)"),
    ("08", "Diagnosis & Staging"),
    ("09", "Treatment"),
    ("10", "Prognosis & Complications"),
]
col1 = topics[:5]
col2 = topics[5:]

for i, (num, topic) in enumerate(col1):
    y = 1.5 + i * 0.98
    add_rect(s, 0.5, y, 0.55, 0.62, MID_BLUE)
    add_tb(s, 0.5, y+0.08, 0.55, 0.5, num, 18, bold=True, color=WHITE, align=PP_ALIGN.CENTER)
    add_rect(s, 1.15, y, 5.5, 0.62, WHITE)
    add_tb(s, 1.3, y+0.1, 5.2, 0.45, topic, 14, color=DARK_BLUE)

for i, (num, topic) in enumerate(col2):
    y = 1.5 + i * 0.98
    add_rect(s, 7.0, y, 0.55, 0.62, MID_BLUE)
    add_tb(s, 7.0, y+0.08, 0.55, 0.5, num, 18, bold=True, color=WHITE, align=PP_ALIGN.CENTER)
    add_rect(s, 7.65, y, 5.3, 0.62, WHITE)
    add_tb(s, 7.8, y+0.1, 5.1, 0.45, topic, 14, color=DARK_BLUE)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 3 – Definition & Introduction
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Definition & Introduction")

add_rect(s, 0.45, 1.45, 12.4, 1.5, WHITE)
add_tb(s, 0.65, 1.6, 12, 1.2,
       "Osteosarcoma is a malignant tumour of bone in which the tumour cells form neoplastic "
       "osteoid, bone, or both. It is the most common primary malignant bone tumour, excluding "
       "haematopoietic tumours (e.g. multiple myeloma).",
       14, color=DARK_GREY, wrap=True)

bullets = [
    "Most common non-haematological primary bone malignancy",
    "Tumour cells directly produce osteoid / bone matrix — the diagnostic hallmark",
    "High-grade intramedullary lesion in 75% of all cases (conventional type)",
    "All osteosarcomas are assumed to have occult micrometastases at time of diagnosis",
    "5-year survival ~70% (no metastasis) with modern chemotherapy + surgery",
]
add_bullet_tf(s, 0.65, 3.15, 12, 3.8, bullets, size=13.5, color=DARK_GREY)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 4 – Epidemiology
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Epidemiology & Risk Factors")

# Left panel - Epidemiology
add_rect(s, 0.45, 1.4, 5.9, 5.6, WHITE)
add_tb(s, 0.65, 1.55, 5.6, 0.5, "Epidemiology", 16, bold=True, color=DARK_BLUE)
epi_bullets = [
    "Incidence: ~4 per million population/year",
    "Bimodal age distribution:",
    "   Peak 1 — Adolescents 10–20 yrs (75% of cases)",
    "   Peak 2 — Older adults (secondary OSA)",
    "Male > Female (ratio 1.35–1.6 : 1)",
    "Peak coincides with adolescent growth spurt",
    "Almost 50%–75% occur near the knee",
    "   (distal femur or proximal tibia)",
]
add_bullet_tf(s, 0.65, 2.1, 5.6, 4.7, epi_bullets, size=12.5, color=DARK_GREY)

# Right panel - Risk factors
add_rect(s, 6.9, 1.4, 6.0, 5.6, WHITE)
add_tb(s, 7.1, 1.55, 5.7, 0.5, "Risk Factors / Associations", 16, bold=True, color=DARK_BLUE)
rf_bullets = [
    ("⚑  ", "Paget disease of bone"),
    ("⚑  ", "Previous radiation therapy"),
    ("⚑  ", "Bone infarcts"),
    ("⚑  ", "Li-Fraumeni syndrome (TP53 germline)"),
    ("⚑  ", "Familial retinoblastoma (RB1 germline)"),
    ("⚑  ", "Rothmund-Thomson syndrome"),
    ("⚑  ", "Fibrous dysplasia"),
    ("⚑  ", "Bone infarction / osteonecrosis"),
    ("⚑  ", "Prior low-grade sarcoma dedifferentiation"),
]
add_bullet_tf(s, 7.1, 2.1, 5.7, 4.7, rf_bullets, size=12.5, color=DARK_GREY)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 5 – WHO Classification
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Classification of Osteosarcoma", "WHO 2013 — Seven Categories")

types = [
    ("1", "Conventional Osteosarcoma", "75% of all cases. High-grade intramedullary. Osteoblastic, chondroblastic or fibroblastic subtypes.", MID_BLUE),
    ("2", "Telangiectatic Osteosarcoma", "Blood-filled cavities. May mimic ABC on X-ray. High-grade.", DARK_BLUE),
    ("3", "Low-Grade Central OSA", "Resembles fibrous dysplasia. Well-differentiated. Better prognosis.", RGBColor(0x27, 0xAE, 0x60)),
    ("4", "Small Cell Osteosarcoma", "Resembles Ewing sarcoma. Small round blue cells + osteoid.", RED_ACCENT),
    ("5", "Parosteal Osteosarcoma", "Surface lesion, posterior distal femur. Low-grade. Best prognosis.", RGBColor(0x84, 0x44, 0xAD)),
    ("6", "Periosteal Osteosarcoma", "Surface, intermediate grade. Chondroblastic variant.", RGBColor(0x16, 0xA0, 0x85)),
    ("7", "High-Grade Surface OSA", "Rare. Surface location but high-grade histology.", RGBColor(0xD3, 0x5A, 0x00)),
]

for i, (num, name, desc, col) in enumerate(types):
    row = i % 4
    col_idx = i // 4
    x = 0.45 + col_idx * 6.55
    y = 1.5 + row * 1.45
    add_rect(s, x, y, 6.1, 1.3, WHITE)
    add_rect(s, x, y, 0.45, 1.3, col)
    add_tb(s, x+0.02, y+0.3, 0.45, 0.7, num, 18, bold=True, color=WHITE, align=PP_ALIGN.CENTER)
    add_tb(s, x+0.6, y+0.07, 5.3, 0.45, name, 13, bold=True, color=col)
    add_tb(s, x+0.6, y+0.52, 5.3, 0.72, desc, 11, color=DARK_GREY, wrap=True)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 6 – Pathogenesis & Molecular Genetics
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Pathogenesis & Molecular Genetics")

add_tb(s, 0.5, 1.35, 12.5, 0.4,
       "Peak incidence during adolescent growth spurt — rapidly proliferating metaphyseal cells prone to mutations",
       13, italic=True, color=MID_GREY)

genes = [
    ("RB1", "Retinoblastoma gene", "Mutated in up to 70% of sporadic OSA\nGermline RB1 mutations increase risk 1000-fold", MID_BLUE),
    ("TP53", "Tumour suppressor", "Mutated in Li-Fraumeni syndrome\nCommon in sporadic tumours", RED_ACCENT),
    ("MDM2/CDK4", "Oncogenes", "Overexpressed in low-grade OSA\nInhibit p53 and RB function respectively", RGBColor(0x27, 0xAE, 0x60)),
    ("CDKN2A", "INK4a locus", "Encodes p16 and p14 tumour suppressors\nInactivated in many OSA", RGBColor(0xD3, 0x5A, 0x00)),
    ("MYC", "Proto-oncogene", "Amplified in up to 50% of cases\nAssociated with particularly poor prognosis", RGBColor(0x84, 0x44, 0xAD)),
]

for i, (gene, role, detail, col) in enumerate(genes):
    x = 0.45 + i * 2.55
    add_rect(s, x, 1.95, 2.3, 4.9, WHITE)
    add_rect(s, x, 1.95, 2.3, 0.65, col)
    add_tb(s, x+0.1, 2.02, 2.1, 0.5, gene, 16, bold=True, color=WHITE)
    add_tb(s, x+0.1, 2.72, 2.1, 0.45, role, 11, bold=True, color=col)
    add_tb(s, x+0.1, 3.2, 2.1, 3.5, detail, 11, color=DARK_GREY, wrap=True)

add_tb(s, 0.5, 7.0, 12.5, 0.38,
       "Note: Complex karyotypes with numerous chromosomal aberrations characterise osteosarcoma",
       12, italic=True, color=MID_GREY)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 7 – Clinical Features
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Clinical Features")

# Symptoms
add_rect(s, 0.45, 1.4, 5.9, 2.75, WHITE)
add_tb(s, 0.65, 1.52, 5.6, 0.42, "Symptoms", 15, bold=True, color=DARK_BLUE)
syms = [
    "Pain — most common presenting complaint",
    "Progressively enlarging painful mass",
    "Night pain (characteristic)",
    "Pathological fracture (first presentation in some)",
    "Neurological symptoms (70% in spinal OSA)",
]
add_bullet_tf(s, 0.65, 2.0, 5.6, 2.0, syms, size=12.5)

# Sites
add_rect(s, 0.45, 4.3, 5.9, 2.75, WHITE)
add_tb(s, 0.65, 4.42, 5.6, 0.42, "Common Sites", 15, bold=True, color=DARK_BLUE)
sites = [
    "Distal femur (most common, ~30%)",
    "Proximal tibia (~25%)",
    "Proximal humerus (~15%)",
    "50%–75% occur near the knee",
    "Rare: ilium, ribs, scapula, jaw, spine",
]
add_bullet_tf(s, 0.65, 4.85, 5.6, 2.0, sites, size=12.5)

# Signs
add_rect(s, 6.9, 1.4, 6.0, 2.75, WHITE)
add_tb(s, 7.1, 1.52, 5.7, 0.42, "Signs on Examination", 15, bold=True, color=DARK_BLUE)
signs = [
    "Hard, tender, fixed mass over metaphysis",
    "Overlying skin warm and erythematous",
    "Dilated superficial veins",
    "Limited adjacent joint movement",
    "Muscle wasting distal to tumour",
    "Lymphadenopathy (uncommon)",
]
add_bullet_tf(s, 7.1, 1.95, 5.7, 2.0, signs, size=12.5)

# Red flags
add_rect(s, 6.9, 4.3, 6.0, 2.75, WHITE)
add_rect(s, 6.9, 4.3, 6.0, 0.5, RED_ACCENT)
add_tb(s, 7.0, 4.37, 5.7, 0.42, "🚩  Red Flag Features", 15, bold=True, color=WHITE)
flags = [
    "Age 10–20 with bone pain + swelling",
    "Pain worse at rest / night",
    "Rapid progression",
    "Raised ALP and LDH (markers of activity)",
]
add_bullet_tf(s, 7.1, 4.9, 5.7, 2.0, flags, size=12.5, color=RED_ACCENT)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 8 – Radiological Features
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Radiological Features")

# X-ray image (Codman triangle)
if imgs[0]["base64"]:
    add_image_from_b64(s, imgs[0]["base64"], 0.45, 1.35, 4.0, 5.8)
add_tb(s, 0.45, 7.1, 4.1, 0.35,
       "X-ray: Distal femur OSA with Codman triangle (arrow)", 10, italic=True, color=MID_GREY)

# Radiology bullet points
add_rect(s, 4.9, 1.35, 8.0, 5.8, WHITE)
add_tb(s, 5.1, 1.5, 7.7, 0.45, "Plain Radiograph Findings", 16, bold=True, color=DARK_BLUE)
rad_bullets = [
    "Metaphyseal location (long bones)",
    "Mixed lytic + sclerotic pattern (most common)",
    "Entirely lytic in 13% of cases",
    "Entirely sclerotic (ivory-like) in some",
    "Cortical destruction with eccentric soft-tissue mass",
    "Periosteal reactions (pathognomonic patterns):",
    "   • Codman triangle — reactive subperiosteal bone",
    "   • Sunburst / spiculated pattern",
    "   • Onion-skin (lamellated) pattern",
    "   • Any combination may be present",
    "",
    "MRI (mandatory for staging):",
    "   • Defines tumour extent in bone & soft tissue",
    "   • Identifies skip metastases",
    "   • Neurovascular involvement",
    "   • Joint involvement",
]
add_bullet_tf(s, 5.1, 2.05, 7.7, 5.0, rad_bullets, size=12)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 9 – Gross & Histopathology
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Pathology — Gross & Histology")

# Gross image
if imgs[1]["base64"]:
    add_image_from_b64(s, imgs[1]["base64"], 0.45, 1.35, 3.8, 4.5)
add_tb(s, 0.45, 5.9, 3.9, 0.5,
       "Gross: Distal femur OSA — cortical disruption,\nhaemorrhage, confined to metaphysis", 10, italic=True, color=MID_GREY, wrap=True)

# Histology image
if imgs[2]["base64"]:
    add_image_from_b64(s, imgs[2]["base64"], 4.55, 1.35, 3.8, 4.5)
add_tb(s, 4.55, 5.9, 3.9, 0.5,
       "Histology: Lancelike osteoid by pleomorphic cells;\nabnormal mitotic figure (arrow)", 10, italic=True, color=MID_GREY, wrap=True)

# Text panel
add_rect(s, 8.75, 1.35, 4.15, 5.75, WHITE)
add_tb(s, 8.95, 1.5, 3.9, 0.42, "Key Pathological Features", 14, bold=True, color=DARK_BLUE)
path_bullets = [
    "Bulky, gritty, tan-white tumour",
    "Areas of haemorrhage and necrosis",
    "Intramedullary spread — replaces marrow",
    "Cortical destruction into soft tissue",
    "Rarely enters adjacent joint",
    "",
    "Histology:",
    "• Cell pleomorphism + hyperchromatic nuclei",
    "• Bizarre tumour giant cells",
    "• Abundant mitoses (incl. abnormal bipolar)",
    "• Fine lacelike or sheet-like osteoid — ESSENTIAL for Dx",
    "• Intravascular invasion common",
]
add_bullet_tf(s, 8.95, 2.0, 3.9, 5.0, path_bullets, size=11.5, color=DARK_GREY)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 10 – Diagnosis & Staging
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Diagnosis & Staging")

# Left
add_rect(s, 0.45, 1.4, 5.9, 5.65, WHITE)
add_tb(s, 0.65, 1.55, 5.6, 0.45, "Investigations", 15, bold=True, color=DARK_BLUE)
inv = [
    ("Lab:", "ALP ↑, LDH ↑ (activity markers); FBC; renal/liver function"),
    ("X-ray:", "First-line; classic periosteal patterns"),
    ("MRI:", "MANDATORY for local staging — whole bone"),
    ("CT chest:", "Screen for pulmonary metastases (commonest site)"),
    ("Bone scan:", "Identify skip metastases, bone metastases"),
    ("PET-CT:", "More sensitive than bone scan in paediatric sarcoma"),
    ("Biopsy:", "DEFINITIVE diagnosis — open or core needle\n(in specialist centre, along tumour axis)"),
]
y = 2.05
for label, text in inv:
    tb = s.shapes.add_textbox(Inches(0.65), Inches(y), Inches(5.6), Inches(0.7))
    tf = tb.text_frame; tf.word_wrap = True
    tf.margin_left=0; tf.margin_right=0; tf.margin_top=0; tf.margin_bottom=0
    p = tf.paragraphs[0]
    r1 = p.add_run(); r1.text = label + " "; r1.font.bold = True
    r1.font.size = Pt(12); r1.font.color.rgb = DARK_BLUE; r1.font.name = "Calibri"
    r2 = p.add_run(); r2.text = text
    r2.font.size = Pt(12); r2.font.color.rgb = DARK_GREY; r2.font.name = "Calibri"
    y += 0.72

# Right
add_rect(s, 6.9, 1.4, 6.0, 5.65, WHITE)
add_tb(s, 7.1, 1.55, 5.7, 0.45, "Staging — Enneking / MSTS System", 15, bold=True, color=DARK_BLUE)
staging = [
    ("Stage IA:", "Low-grade, intracompartmental (T1)"),
    ("Stage IB:", "Low-grade, extracompartmental (T2)"),
    ("Stage IIA:", "High-grade, intracompartmental (T1)"),
    ("Stage IIB:", "High-grade, extracompartmental (T2) — MOST OSA"),
    ("Stage III:", "Any grade with regional or distant metastasis"),
]
y2 = 2.05
for label, text in staging:
    tb = s.shapes.add_textbox(Inches(7.1), Inches(y2), Inches(5.7), Inches(0.6))
    tf = tb.text_frame; tf.word_wrap = True
    tf.margin_left=0; tf.margin_right=0; tf.margin_top=0; tf.margin_bottom=0
    p = tf.paragraphs[0]
    r1 = p.add_run(); r1.text = label + " "
    r1.font.bold = True; r1.font.size = Pt(12); r1.font.color.rgb = MID_BLUE; r1.font.name = "Calibri"
    r2 = p.add_run(); r2.text = text
    r2.font.size = Pt(12); r2.font.color.rgb = DARK_GREY; r2.font.name = "Calibri"
    y2 += 0.68

add_rect(s, 6.9, 5.4, 6.0, 1.65, LIGHT_BLUE)
add_tb(s, 7.1, 5.5, 5.7, 1.4,
       "Biopsy is the gold standard for diagnosis.\n"
       "Biopsy tract must be excised en bloc with the tumour.\n"
       "Always perform in a specialist sarcoma centre.",
       12, color=DARK_BLUE, wrap=True, italic=True)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 11 – Treatment
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Treatment")

# Timeline bar
steps = [
    ("1\nNeoadjuvant\nChemo", MID_BLUE, "MAP protocol\n(Methotrexate,\nAdriaymcin, Cisplatin)\n8–12 weeks"),
    ("2\nSurgery", DARK_BLUE, "Limb-salvage surgery\n(preferred) or\namputation if\nneurovascular involved"),
    ("3\nHistological\nResponse", RGBColor(0x27, 0xAE, 0x60), "Assess tumour necrosis\n≥90% necrosis = good\nresponse (favourable)"),
    ("4\nAdjuvant\nChemo", MID_BLUE, "Continue MAP\n(± ifosfamide if\npoor responder)\n~12–18 weeks"),
]

for i, (step, col, detail) in enumerate(steps):
    x = 0.45 + i * 3.2
    add_rect(s, x, 1.4, 2.9, 1.0, col)
    add_tb(s, x, 1.4, 2.9, 1.0, step, 13, bold=True, color=WHITE, align=PP_ALIGN.CENTER, wrap=True)
    add_rect(s, x, 2.5, 2.9, 2.2, WHITE)
    add_tb(s, x+0.15, 2.6, 2.6, 2.0, detail, 12, color=DARK_GREY, wrap=True)
    if i < 3:
        add_tb(s, x+2.95, 1.75, 0.25, 0.5, "→", 22, bold=True, color=DARK_BLUE, align=PP_ALIGN.CENTER)

# Surgery details
add_rect(s, 0.45, 4.9, 12.45, 2.3, WHITE)
add_tb(s, 0.65, 5.05, 12, 0.4, "Surgical Principles", 15, bold=True, color=DARK_BLUE)
surg = [
    "Limb-salvage surgery (LSS) is now possible in ~85–90% of cases due to effective neoadjuvant chemotherapy",
    "Wide margin excision (en bloc) with reconstruction: endoprosthesis, allograft, or rotationplasty",
    "Amputation: reserved for unresectable cases or where LSS compromises oncological outcome",
    "Radiation therapy: limited role (OSA is relatively radio-resistant); used for unresectable spinal or pelvic lesions",
]
add_bullet_tf(s, 0.65, 5.5, 12, 1.6, surg, size=12)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 12 – Prognosis & Complications
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Prognosis & Complications")

# Prognosis boxes
prog_data = [
    ("~70%", "5-year survival\n(no metastasis)", MID_BLUE),
    ("<30%", "5-year survival\n(with metastasis)", RED_ACCENT),
    ("≥90%", "Tumour necrosis\n= good response", RGBColor(0x27, 0xAE, 0x60)),
    ("~20%", "5-year survival\nspinal OSA", RGBColor(0xD3, 0x5A, 0x00)),
]
for i, (val, label, col) in enumerate(prog_data):
    x = 0.45 + i * 3.22
    add_rect(s, x, 1.4, 2.9, 1.5, col)
    add_tb(s, x, 1.55, 2.9, 0.85, val, 38, bold=True, color=WHITE, align=PP_ALIGN.CENTER)
    add_tb(s, x, 2.3, 2.9, 0.6, label, 12, color=WHITE, align=PP_ALIGN.CENTER, wrap=True)

# Prognostic factors
add_rect(s, 0.45, 3.1, 5.9, 4.05, WHITE)
add_tb(s, 0.65, 3.25, 5.6, 0.42, "Favourable Prognostic Factors", 14, bold=True, color=RGBColor(0x27, 0xAE, 0x60))
fav = [
    "Age < 40 years",
    "Distal extremity location",
    "No metastasis at diagnosis",
    "≥90% histological necrosis post-chemo",
    "Complete surgical resection (wide margins)",
    "Low ALP at diagnosis",
]
add_bullet_tf(s, 0.65, 3.75, 5.6, 3.2, fav, size=12.5, color=DARK_GREY)

add_rect(s, 6.9, 3.1, 6.0, 4.05, WHITE)
add_tb(s, 7.1, 3.25, 5.7, 0.42, "Complications & Metastases", 14, bold=True, color=RED_ACCENT)
comp = [
    "Haematogenous spread — lungs (most common)",
    "Bone metastases, brain, other sites",
    "Skip metastases (same bone, across joint)",
    "Pathological fracture",
    "Local recurrence after surgery",
    "Chemotherapy toxicity (cardiotoxicity, nephrotoxicity)",
    "Prosthesis failure / aseptic loosening",
    "Poor prognosis if clinically evident metastasis at Dx",
]
add_bullet_tf(s, 7.1, 3.75, 5.7, 3.2, comp, size=12.5, color=DARK_GREY)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 13 – Differential Diagnosis
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, LIGHT_GREY)
slide_header(s, "Differential Diagnosis")

diffs = [
    ("Ewing Sarcoma", "Age 10–20; diaphysis; 'onion skin' periosteal reaction;\nsmall round blue cells on histology; no osteoid", MID_BLUE),
    ("Chondrosarcoma", "Adults >40; lytic with rings & arcs calcification;\ncartilage matrix; low to intermediate grade", DARK_BLUE),
    ("Giant Cell Tumour", "Age 20–40; epiphysis; soap-bubble lytic lesion;\nno matrix; osteoclast-like giant cells", RGBColor(0x27, 0xAE, 0x60)),
    ("Osteoid Osteoma", "Age 10–30; diaphysis; nocturnal pain relieved by aspirin;\nnidus <2 cm; reactive sclerosis; benign", RGBColor(0xD3, 0x5A, 0x00)),
    ("Metastatic Bone Disease", "Adults >40; multiple lesions; known primary elsewhere;\nlytic (most) or sclerotic (prostate, breast)", RED_ACCENT),
    ("Aneurysmal Bone Cyst", "Any age; expansile lytic; fluid-fluid levels on MRI;\ncan mimic telangiectatic OSA; benign", RGBColor(0x84, 0x44, 0xAD)),
]

for i, (name, desc, col) in enumerate(diffs):
    row = i % 3
    ci = i // 3
    x = 0.45 + ci * 6.55
    y = 1.4 + row * 1.9
    add_rect(s, x, y, 6.1, 1.75, WHITE)
    add_rect(s, x, y, 0.1, 1.75, col)
    add_tb(s, x+0.25, y+0.12, 5.7, 0.45, name, 14, bold=True, color=col)
    add_tb(s, x+0.25, y+0.57, 5.7, 1.1, desc, 12, color=DARK_GREY, wrap=True)

# ═══════════════════════════════════════════════════════════════════════════
# SLIDE 14 – Summary
# ═══════════════════════════════════════════════════════════════════════════
s = prs.slides.add_slide(blank)
add_rect(s, 0, 0, 13.333, 7.5, DARK_BLUE)
add_rect(s, 0, 5.5, 13.333, 2.0, MID_BLUE)
add_rect(s, 0.55, 1.6, 0.07, 3.5, RED_ACCENT)

add_tb(s, 0.9, 1.35, 11.5, 0.8, "Key Take-Home Points", 30, bold=True,
       color=WHITE, align=PP_ALIGN.LEFT)

summary = [
    "Most common primary malignant bone tumour (excl. haematological)",
    "Bimodal age: adolescent peak (10–20 yrs) + older adults (secondary OSA)",
    "Metaphysis of long bones around the knee in 50–75% of cases",
    "Diagnostic hallmark: malignant cells producing osteoid / bone matrix",
    "Classic X-ray: mixed lytic-sclerotic, Codman triangle, sunburst periosteal reaction",
    "Treatment: Neoadjuvant chemo → Limb-salvage surgery → Adjuvant chemo (MAP)",
    "5-year survival ~70% without metastasis; <30% with metastasis",
    "Lungs are the most common site of distant metastasis",
]
for i, point in enumerate(summary):
    add_tb(s, 1.05, 2.35 + i * 0.59, 11.5, 0.55,
           f"  ✓  {point}", 13, color=WHITE if i % 2 == 0 else RGBColor(0xAD, 0xD4, 0xF0),
           wrap=True)

add_tb(s, 0.9, 5.7, 9, 0.5, "Sources: Robbins & Kumar Basic Pathology | Grainger & Allison Diagnostic Radiology | Campbell's Operative Orthopaedics 2026", 11,
       color=WHITE, align=PP_ALIGN.LEFT, italic=True)

# ── Save ──────────────────────────────────────────────────────────────────
out = "/tmp/workspace/osteosarcoma-seminar/Osteosarcoma_Seminar_MBBS.pptx"
prs.save(out)
print(f"Saved: {out}")

Successfully replaced 1 instance(s) of the string in '/tmp/workspace/osteosarcoma-seminar/build_pptx.py'

Successfully replaced 1 instance(s) of the string in '/tmp/workspace/osteosarcoma-seminar/build_pptx.py'