The week in GI, filtered to what is worth knowing.

Confirm ulcerative colitis with endoscopy showing continuous colonic inflammation from the rectum plus histology, after excluding infection with two-step CDI testing. Treat mild-moderate disease with 5-ASA by route and extent (suppository for proctitis, enema for left-sided, oral plus rectal for extensive); if 5-ASA fails, treat as moderate-to-severe rather than stepping up gradually. Position advanced therapy by EFFICACY TIER, not by an anti-TNF-first rule. Advanced-therapy-naive, higher efficacy: infliximab, vedolizumab, ozanimod, etrasimod, upadacitinib, risankizumab, guselkumab; intermediate: golimumab, ustekinumab, tofacitinib, filgotinib, mirikizumab; lower: adalimumab. PREVIOUSLY TNF-EXPOSED, higher efficacy: tofacitinib, upadacitinib, ustekinumab; lower: adalimumab, vedolizumab, ozanimod, etrasimod - S1P modulators are weakest in exactly this group. Do not cycle within the anti-TNF class after primary non-response; switch mechanism. Apply treat-to-target (STRIDE-II) to endoscopic improvement (MES 0-1). Screen for acute severe UC by Truelove-Witts and admit for IV steroids. Begin CRC surveillance 8-10 years after DIAGNOSIS for extensive or left-sided disease; isolated proctitis follows average-risk screening.

Perform a diagnostic paracentesis in all patients with new-onset ascites that is accessible for sampling, and in any patient with cirrhosis and ascites who is emergently admitted to the hospital for any reason, to rule out SBP even in the absence of signs or symptoms of infection (up to one-third of SBP is asymptomatic). Also perform it whenever a patient with ascites develops signs, symptoms, or laboratory abnormalities suggestive of infection, and in patients with tense ascites and AKI to exclude SBP as the cause. LVP/paracentesis is safe even with coagulopathy: elevated prothrombin time (INR >1.5) or thrombocytopenia (platelets <50×10^9/L) is not a contraindication, and routine transfusion of clotting factors or platelets is NOT recommended, with possible exceptions in disseminated intravascular coagulation, or uremia with thrombocytopenia (where desmopressin may be considered, especially with prior bleeding). The initial laboratory investigation of ascitic fluid should include ascitic fluid neutrophil (PMN) count, ascitic fluid total protein, and ascitic fluid albumin plus a simultaneous serum albumin to calculate the serum-ascites albumin gradient (SAAG). Ascitic fluid culture is not part of that routine initial panel but should be obtained whenever SBP is being evaluated (routinely for inpatients and when secondary bacterial peritonitis is suspected); when cultured, inoculate ≥10 mL of fluid into aerobic and anaerobic blood culture bottles at the bedside, before the first dose of antibiotics. Amylase, cytology, glucose, LDH, and mycobacterial (AFB/ADA) or triglyceride testing are not routinely indicated and should be guided by clinical context (e.g., suspected pancreatic, malignant, secondary-bacterial, tuberculous, or chylous ascites). Interpret SAAG to classify the cause: SAAG ≥1.1 g/dL indicates portal hypertension (~97% accuracy), while SAAG <1.1 g/dL points to non-portal-hypertensive (peritoneal) causes such as peritoneal carcinomatosis, tuberculous peritonitis, nephrotic syndrome, or pancreatic/chylous ascites. Within high-SAAG ascites, use ascitic total protein to distinguish the source: total protein <2.5 g/dL suggests cirrhosis (also massive liver metastasis, late Budd-Chiari), whereas total protein ≥2.5 g/dL suggests a postsinusoidal/cardiac source (heart failure, portal vein thrombosis, sinusoidal obstruction syndrome, early Budd-Chiari), directing workup toward right-heart catheterization, hepatic-vein Doppler, and echocardiography. The diagnosis of SBP is established with an ascitic fluid absolute neutrophil (PMN) count >250/mm^3 (with or without a positive culture, and in the absence of a surgically treatable intra-abdominal source of infection). Spontaneous bacterial empyema (SBE) is the analogous spontaneous infection of a pre-existing hepatic hydrothorax, not a parapneumonic effusion, and is established by the same pleural fluid PMN count >250/mm^3, but only after pneumonia and other secondary causes of the effusion have been excluded; despite the term "empyema" it is managed as an infection with antibiotics, and a chest tube should NOT be placed.

Scope: this primary document covers portal hypertension and varices in cirrhosis ONLY. Everything else in the current entry (HCC surveillance, transplant referral, nutrition, vaccination, ascites/SBP/HE/HRS/ACLF management) belongs to the separate documents listed as supporting sources and must not be attributed to this guidance. Risk-stratify compensated advanced chronic liver disease (cACLD) with non-invasive tests before defaulting to endoscopy. By vibration-controlled transient elastography, liver stiffness (LSM) under 10 kPa excludes cACLD, 10 to 14.9 kPa is suggestive and warrants further testing, and 15 kPa or above rules cACLD in. Clinically significant portal hypertension (CSPH, i.e. HVPG above 10 mm Hg) is ruled in non-invasively by LSM at or above 25 kPa alone, or LSM 20 to 24.9 kPa with platelets below 150 x 10^9/L, or LSM 15 to 19.9 kPa with platelets below 110 x 10^9/L. Where spleen stiffness measurement is available, values below 21 kPa argue against CSPH and above 50 kPa argue for it. These non-invasive cutoffs are conditional probability thresholds, not universal diagnostic rules, and hold only when their derivation conditions are met. They were established in compensated advanced chronic liver disease of viral (treated or untreated HBV/HCV), alcohol-related, or non-obese (BMI under 30) MASLD etiology, and are not validated for and should not be applied to cholestatic (e.g. PBC, PSC), vascular, or other etiologies, in which different cutoffs apply. They also require a technically reliable liver stiffness measurement by vibration-controlled transient elastography, fasting, at least 10 valid acquisitions, and an IQR/median ratio below 30 percent, and are confounded by factors that raise liver stiffness independently of fibrosis, including ALT flares or acute hepatitis, extrahepatic cholestasis, hepatic congestion (right heart failure), and recent food intake; Baveno VII recommends confirming an LSM-based rule-in on a repeat measurement obtained on a separate day. The spleen stiffness cutoffs apply only to a dedicated 100 Hz spleen module, carry a higher technical failure rate, are machine-dependent, and are not interchangeable with values obtained using the standard 50 Hz liver probe. Where these population or acquisition conditions are not met, the numeric thresholds do not reliably classify CSPH, and HVPG measurement or endoscopy is required rather than a non-invasive determination. Preserve the gray zone: LSM 15 to 25 kPa without a qualifying platelet count is explicitly indeterminate (roughly 40 to 60 percent of cACLD patients) and calls for HVPG and/or endoscopy rather than assumption in either direction. Once CSPH is established, offer a non-selective beta-blocker, carvedilol preferred at 12.5 mg/day, to lower portal pressure and prevent FIRST decompensation, on the strength of PREDESCI. This is a treat-the-pressure strategy rather than a treat-the-varices one: because a patient already on NSBB for CSPH would not have management changed by the finding, screening endoscopy becomes unnecessary for most such patients. Screening EGD is still indicated where elastography is unavailable, technically difficult, or gives discordant results, and for patients who do not meet non-invasive criteria. Standard beta-blocker contraindications (severe asthma, advanced heart block, bradyarrhythmias) continue to apply. For varices needing treatment managed endoscopically, band ligation is repeated every 2 to 4 weeks until obliteration, with surveillance endoscopy at 6 months and annually thereafter. Gastric/cardiofundal varices are treated with cyanoacrylate injection or EUS-guided coil embolization rather than banding, with hepatology involvement. In acute variceal hemorrhage, consider preemptive (early) TIPS within 24 to 72 hours in high-risk patients: Child-Turcotte-Pugh class B with score above 7 and active bleeding at endoscopy despite vasoactive therapy, or CTP class C with score 10 to 13. Keep the document's own qualifier: there is no established universal MELD or CTP cutoff at which TIPS is mandatory, and the decision is case-by-case with hepatology consultation. Statin use in compensated cirrhosis is NOT a recommendation traceable to this document; the current entry's statin sentence needs its own source or removal. Verification note: the Hepatology full text is paywalled (402/403 on direct fetch). Document identity, title, year, volume, pages and authorship were confirmed via Crossref and PubMed metadata; the numeric thresholds above were corroborated from two independent peer-reviewed commentaries on this specific guidance and match the Baveno VII rule-of-five, but were not read verbatim from the primary PDF. A human should confirm the exact guidance-statement wording against the full text before this entry is treated as quotation-grade.

Scope: this primary document covers portal hypertension and varices in cirrhosis ONLY. Everything else in the current entry (HCC surveillance, transplant referral, nutrition, vaccination, ascites/SBP/HE/HRS/ACLF management) belongs to the separate documents listed as supporting sources and must not be attributed to this guidance. Risk-stratify compensated advanced chronic liver disease (cACLD) with non-invasive tests before defaulting to endoscopy. By vibration-controlled transient elastography, liver stiffness (LSM) under 10 kPa excludes cACLD, 10 to 14.9 kPa is suggestive and warrants further testing, and 15 kPa or above rules cACLD in. Clinically significant portal hypertension (CSPH, i.e. HVPG above 10 mm Hg) is ruled in non-invasively by LSM at or above 25 kPa alone, or LSM 20 to 24.9 kPa with platelets below 150 x 10^9/L, or LSM 15 to 19.9 kPa with platelets below 110 x 10^9/L. Where spleen stiffness measurement is available, values below 21 kPa argue against CSPH and above 50 kPa argue for it. These non-invasive cutoffs are conditional probability thresholds, not universal diagnostic rules, and hold only when their derivation conditions are met. They were established in compensated advanced chronic liver disease of viral (treated or untreated HBV/HCV), alcohol-related, or non-obese (BMI under 30) MASLD etiology, and are not validated for and should not be applied to cholestatic (e.g. PBC, PSC), vascular, or other etiologies, in which different cutoffs apply. They also require a technically reliable liver stiffness measurement by vibration-controlled transient elastography, fasting, at least 10 valid acquisitions, and an IQR/median ratio below 30 percent, and are confounded by factors that raise liver stiffness independently of fibrosis, including ALT flares or acute hepatitis, extrahepatic cholestasis, hepatic congestion (right heart failure), and recent food intake; Baveno VII recommends confirming an LSM-based rule-in on a repeat measurement obtained on a separate day. The spleen stiffness cutoffs apply only to a dedicated 100 Hz spleen module, carry a higher technical failure rate, are machine-dependent, and are not interchangeable with values obtained using the standard 50 Hz liver probe. Where these population or acquisition conditions are not met, the numeric thresholds do not reliably classify CSPH, and HVPG measurement or endoscopy is required rather than a non-invasive determination. Preserve the gray zone: LSM 15 to 25 kPa without a qualifying platelet count is explicitly indeterminate (roughly 40 to 60 percent of cACLD patients) and calls for HVPG and/or endoscopy rather than assumption in either direction. Once CSPH is established, offer a non-selective beta-blocker, carvedilol preferred at 12.5 mg/day, to lower portal pressure and prevent FIRST decompensation, on the strength of PREDESCI. This is a treat-the-pressure strategy rather than a treat-the-varices one: because a patient already on NSBB for CSPH would not have management changed by the finding, screening endoscopy becomes unnecessary for most such patients. Screening EGD is still indicated where elastography is unavailable, technically difficult, or gives discordant results, and for patients who do not meet non-invasive criteria. Standard beta-blocker contraindications (severe asthma, advanced heart block, bradyarrhythmias) continue to apply. For varices needing treatment managed endoscopically, band ligation is repeated every 2 to 4 weeks until obliteration, with surveillance endoscopy at 6 months and annually thereafter. Gastric/cardiofundal varices are treated with cyanoacrylate injection or EUS-guided coil embolization rather than banding, with hepatology involvement. In acute variceal hemorrhage, consider preemptive (early) TIPS within 24 to 72 hours in high-risk patients: Child-Turcotte-Pugh class B with score above 7 and active bleeding at endoscopy despite vasoactive therapy, or CTP class C with score 10 to 13. Keep the document's own qualifier: there is no established universal MELD or CTP cutoff at which TIPS is mandatory, and the decision is case-by-case with hepatology consultation. Statin use in compensated cirrhosis is NOT a recommendation traceable to this document; the current entry's statin sentence needs its own source or removal. Verification note: the Hepatology full text is paywalled (402/403 on direct fetch). Document identity, title, year, volume, pages and authorship were confirmed via Crossref and PubMed metadata; the numeric thresholds above were corroborated from two independent peer-reviewed commentaries on this specific guidance and match the Baveno VII rule-of-five, but were not read verbatim from the primary PDF. A human should confirm the exact guidance-statement wording against the full text before this entry is treated as quotation-grade.

Perform HCC surveillance in adults with cirrhosis of any etiology who are able to benefit from HCC treatment, accordingly, do not enroll patients with Child-Turcotte-Pugh class C cirrhosis unless they are candidates for (or listed for) liver transplantation, and do not surveil those with life-limiting comorbidity (life expectancy <1-2 years) that transplantation cannot remedy, and in select non-cirrhotic chronic HBV carriers at elevated risk (e.g., active hepatitis/high viral load, family history of HCC, African/African-American ancestry, Asian men >40 and Asian women >50; AASLD references risk scores such as PAGE-B but stops short of mandating a specific numeric cutoff). Surveillance is abdominal ultrasound PLUS serum AFP at ~6-month (semiannual) intervals, a change from prior AASLD guidance, which had left AFP optional (Strong recommendation). Pursue diagnostic evaluation for an AFP ≥20 ng/mL or for a rising AFP, defined as a doubling of AFP or an increase on two consecutive tests even if the absolute value is <20 ng/mL. Any suspicious ultrasound observation ≥1 cm, or an inadequate/limited ultrasound, should be evaluated with multiphasic contrast-enhanced CT or MRI interpreted with CT/MRI LI-RADS. AASLD 2023 stages confirmed HCC using the BCLC framework and links stage to therapy (curative resection, ablation, or transplantation within Milan criteria for very-early/early disease; TACE and other locoregional therapy for intermediate disease; systemic therapy for advanced disease). Its first-line systemic options for advanced HCC with preserved liver function (Child-Pugh A) are atezolizumab+bevacizumab OR durvalumab+tremelimumab (STRIDE) as preferred, with sorafenib, lenvatinib, or durvalumab monotherapy as alternatives when the preferred combinations are contraindicated. After potentially curative therapy, continue surveillance for recurrence given high recurrence risk (AASLD monitors with cross-sectional contrast-enhanced imaging, not ultrasound, in the early post-treatment period).

Dose and select GI/hepatology drugs by GFR, and treat imaging contrast as a procedure-specific decision rather than a contraindication. KDIGO 2024 Chapter 4 gives practice points, not graded recommendations. Nephrotoxin stewardship (PP 4.1.1-4.1.3): people with CKD may be more susceptible to nephrotoxic effects, always weigh benefits versus potential harms before prescribing; monitor eGFR, electrolytes and drug levels in those on medications with narrow therapeutic windows, potential adverse effects or nephrotoxicity; review and limit OTC medicines and dietary/herbal remedies. The guideline names two GI-relevant nephrotoxins in Table 31: NSAIDs (reduced prostaglandin-dependent kidney blood flow, acute interstitial nephritis, nephrotic syndrome; alternative = acetaminophen) and proton pump inhibitors (AKI and CKD via tubulointerstitial/acute interstitial nephritis; alternative = H2-receptor antagonists). On NSAIDs the guideline is deliberately conditional, not prohibitive: indiscriminate chronic OTC NSAID use is associated with higher risk of kidney failure and should be discouraged, but 'judicious NSAID use, under careful supervision of a nephrologist, may be preferred to other pain medications such as opioids.' For acute gout specifically (PP 3.14.3), low-dose colchicine or intra-articular/oral glucocorticoids are preferable to NSAIDs (colchicine dose adjustment considered at CKD G5; e.g. prednisolone 30 mg orally for 3-5 days as an alternative). Dosing (PP 4.2.1-4.2.5): consider GFR for renally cleared drugs; validated eGFRcr equations suffice for most drug dosing; use creatinine+cystatin C or measured GFR where more accuracy is needed (narrow therapeutic range, drug toxicity, unreliable eGFRcr); use non-BSA-indexed eGFR at extremes of body weight; adapt dosing when GFR or volume of distribution is not at steady state. Stewardship (PP 4.3.1-4.3.3): perform medication review periodically and at transitions of care; consider planned discontinuation of metformin, ACEi, ARBs and SGLT2i 48-72 hours before elective surgery, but note that failure to restart is itself a source of harm, so a clear documented restart plan must be communicated (PP 4.3.2). Anticoagulation (Rec 3.16.1, 1C): NOACs are recommended in preference to vitamin K antagonists for thromboprophylaxis in atrial fibrillation in CKD G1-G4; NOAC dose adjustment for GFR is required, with caution at G4-G5 (PP 3.16.2); duration of pre-procedure NOAC hold depends on procedural bleeding risk, the specific NOAC, and GFR (PP 3.16.3, Figure 44). Imaging (PP 4.4.1): image on general-population indications, with individualized risk/benefit, the guideline explicitly warns against harm from delayed, omitted or suboptimal imaging driven by fear of contrast. Iodinated contrast: assess AKI risk with validated tools for intra-arterial cardiac procedures (PP 4.4.1.1); manage intravenous contrast per radiology-society consensus statements in people with AKI or GFR <60 (CKD G3a-G5) undergoing elective investigation (PP 4.4.1.2), which the Work Group summarizes as, use low- or iso-osmolar media at the minimum diagnostic dose; withdraw nonessential potentially nephrotoxic medications (NSAIDs, diuretics, aminoglycosides, amphotericin, platins, zoledronate, methotrexate) in people with AKI or eGFR <30 for 24-48 hours before and 48 hours after contrast; at eGFR >30 without AKI metformin need not be stopped and no post-procedure GFR testing is needed, whereas with AKI or eGFR ≤30 metformin should be stopped at or before injection and not restarted for at least 48 hours and only if GFR is stable and use has been reassessed; consider withholding RAASi ≥48 hours before elective contrast-enhanced CT in at-risk people; avoid dehydration in non-dialysis people with eGFR <30 or AKI receiving IV contrast; N-acetylcysteine, ascorbic acid, furosemide, dopamine, fenoldopam and calcium channel blockers have not shown consistent benefit, and prophylactic peri-contrast hemodialysis is potentially harmful and is not recommended. Gadolinium (PP 4.4.2.1): for GFR <30 (CKD G4-G5) requiring gadolinium, preferentially offer ACR group II and III agents, greatest NSF risk is in AKI, KRT and CKD G4-G5, most unconfounded cases involved group I agents, and no case has been reported since 2012.

In patients with classic GERD symptoms (heartburn, regurgitation) and no alarm symptoms, give an 8-week trial of empiric PPI ONCE DAILY before a meal (strong, moderate evidence); administer PPI 30-60 minutes before a meal rather than at bedtime (strong, moderate). Endoscopy is the FIRST test in patients presenting with dysphagia or other alarm symptoms, the guideline names weight loss and GI bleeding, and also in patients with multiple risk factors for Barrett's esophagus (strong, low). If classic symptoms respond to the 8-week trial, attempt to discontinue the PPI (conditional, low); for patients without erosive esophagitis or Barrett's whose symptoms resolved, an attempt at discontinuation should be made; patients requiring maintenance should take the lowest effective dose. If symptoms do not respond adequately to the 8-week trial, or return on discontinuation, perform diagnostic endoscopy, ideally after PPIs are stopped for 2-4 weeks (strong, low). Where GERD is suspected but unclear and endoscopy shows no objective evidence, perform reflux monitoring OFF therapy to establish the diagnosis (strong, low); conversely, do NOT perform off-therapy reflux monitoring solely as a diagnostic test in patients already known to have LA grade C or D esophagitis or long-segment Barrett's, because the diagnosis is already established (strong, low). In refractory GERD, optimize PPI therapy first (strong, moderate); then pH monitoring OFF PPIs if GERD was not previously established by pH study, long-segment Barrett's, or LA grade C/D esophagitis, versus impedance-pH ON PPIs where GERD is already established but symptoms persist on twice-daily PPI (both conditional, low). Before antireflux surgery or endoscopic therapy, HRM is recommended to rule out achalasia and absent contractility, with provocative testing (e.g., multiple rapid swallows) to identify contractile reserve in ineffective esophageal motility; on-therapy reflux monitoring is suggested before intervention in patients with prior objective GERD findings who remain symptomatic. Antireflux surgery by an experienced surgeon is an option for patients with OBJECTIVE evidence of GERD, with severe esophagitis (LA C/D), large hiatal hernia, or persistent troublesome symptoms benefiting most; TIF is suggested only for troublesome regurgitation or heartburn in patients who do not wish to undergo antireflux surgery and who are WITHOUT severe esophagitis (LA C/D) or hiatal hernia >2 cm. For extraesophageal symptoms WITHOUT typical GERD symptoms, perform reflux testing BEFORE starting PPI therapy (strong, moderate); with concomitant typical symptoms, consider twice-daily PPI for 8-12 weeks before further testing (conditional, low). Lifestyle: weight loss in overweight/obese patients (strong, moderate); avoid meals within 2-3 hours of bedtime, avoid tobacco, avoid trigger foods, and elevate the head of the bed for nighttime symptoms (all conditional, low).

Stage cholangiocarcinoma with multiphasic MRI/MRCP plus CT chest/abdomen/pelvis, obtain tissue via ERCP brush cytology with FISH, cholangioscopy-directed biopsy, or EUS-FNA (avoiding primary perihilar tumor sampling if liver transplant is being considered due to theoretical seeding risk), and manage through a multidisciplinary tumor board with treatment stratified by tumor location: resect (anatomic hepatectomy for intrahepatic, hemihepatectomy with bile duct resection for perihilar, Whipple for distal) with adjuvant capecitabine (category 1, per BILCAP) when resectable. For unresectable or metastatic disease, first-line systemic therapy is gemcitabine plus cisplatin combined with a PD-L1/PD-1 checkpoint inhibitor, either durvalumab (TOPAZ-1) or pembrolizumab (KEYNOTE-966), both listed as category 1 preferred options by NCCN; comprehensive molecular/genomic profiling is recommended for all unresectable/metastatic candidates for systemic therapy to identify actionable targets (e.g., FGFR2 fusions, IDH1 mutations, HER2 amplification) for later-line therapy. Selected unresectable early-stage perihilar tumors may undergo neoadjuvant chemoradiation and liver transplant per the Mayo Clinic protocol at experienced centers, and jaundice or cholangitis is managed with biliary drainage (ERCP preferred over PTC).

Evaluate suspected pancreatic adenocarcinoma with a dedicated pancreatic-protocol (multiphasic, contrast-enhanced) CT of the abdomen as the preferred imaging for diagnosis and staging (MRI/MRCP is an acceptable alternative); PET/CT may be added after, never as a substitute for, protocol CT/MRI in higher-risk situations. Obtain tissue confirmation, preferably by EUS-guided FNA/FNB with core biopsy where feasible; biopsy is generally required before neoadjuvant or palliative therapy, whereas patients with clearly resectable disease may proceed to surgery without preoperative biopsy. Measure baseline CA 19-9 (interpret after biliary decompression / with normal bilirubin; it can be falsely low in Lewis-antigen-negative patients and falsely elevated with cholestasis). Offer GERMLINE testing for inherited susceptibility to EVERY patient with confirmed pancreatic adenocarcinoma regardless of family history, using a comprehensive multi-gene panel; separately, for patients with locally advanced or metastatic disease who are candidates for anti-cancer therapy, obtain TUMOR/SOMATIC molecular profiling (broad NGS; RNA-based assays improve fusion detection; cell-free/liquid biopsy if tissue is insufficient). Assign resectability, resectable, borderline resectable, locally advanced (unresectable), or metastatic, by multidisciplinary review, ideally at a high-volume center. RESECTABLE disease: upfront resection (pancreaticoduodenectomy for head/uncinate; distal pancreatectomy ± splenectomy for body/tail) followed by adjuvant therapy, or neoadjuvant therapy in selected high-risk patients; preferred adjuvant regimen is modified FOLFIRINOX in patients able to tolerate it, ideally begun within 12 weeks of surgery, for a total of ~6 months of perioperative systemic therapy. BORDERLINE RESECTABLE and LOCALLY ADVANCED disease: neoadjuvant systemic therapy, with or without subsequent chemoradiation/RT, then restaging and reassessment for resection. METASTATIC disease by performance status: good PS (ECOG 0-1, adequate organ function), first-line FOLFIRINOX/modified FOLFIRINOX, NALIRIFOX, or gemcitabine + nab-paclitaxel, with clinical trial preferred; after ~4-6 months without progression, transition to a maintenance strategy (including olaparib for germline BRCA1/2 carriers who have NOT progressed on first-line platinum-based therapy), a clinical trial, or a treatment break; intermediate PS (ECOG 2), clinical trial preferred, otherwise less-intensive first-line therapy (preferred gemcitabine + nab-paclitaxel [category 1], or single-agent gemcitabine or capecitabine, or FOLFOX/FOLFIRI/CapeOx), since the intensive triplets (FOLFIRINOX/modified FOLFIRINOX and NALIRIFOX) are limited to ECOG 0-1 and are generally not appropriate at ECOG 2, with the same maintenance-versus-progression reassessment applied after ~4-6 months without progression; poor PS (ECOG 3-4), palliative/best supportive care, with single-agent chemotherapy or a biomarker-directed agent only if clinically appropriate. Treat by molecular target wherever an actionable alteration is present, across lines: pembrolizumab (or dostarlimab) for MSI-H/dMMR and pembrolizumab for TMB-H, olaparib for germline BRCA1/2 (maintenance after platinum), larotrectinib/entrectinib for NTRK fusions, selpercatinib for RET fusions, dabrafenib + trametinib for BRAF V600E, erdafitinib for FGFR2 alterations, fam-trastuzumab deruxtecan for HER2-positive disease, zenocutuzumab for NRG1 fusion-positive disease, and (in recent NCCN versions) KRAS G12C inhibitors (adagrasib/sotorasib), several of these after prior systemic therapy. Integrate supportive care throughout: relieve obstructive jaundice with an endoscopic self-expanding metal stent in non-operative or neoadjuvant pathways, but do NOT perform routine preoperative biliary drainage before planned upfront resection; treat exocrine insufficiency with pancreatic enzyme replacement titrated to symptoms (common starting dose ~40,000-50,000 units of lipase with main meals, roughly half with snacks, increased as needed); manage refractory pain with celiac plexus neurolysis or palliative RT; relieve gastric outlet obstruction with surgical or EUS-guided gastrojejunostomy (or enteral stenting); and provide VTE prophylaxis, nutritional support, and early palliative care.

Evaluate suspected pancreatic adenocarcinoma with a dedicated pancreatic-protocol (multiphasic, contrast-enhanced) CT of the abdomen as the preferred imaging for diagnosis and staging (MRI/MRCP is an acceptable alternative); PET/CT may be added after, never as a substitute for, protocol CT/MRI in higher-risk situations. Obtain tissue confirmation, preferably by EUS-guided FNA/FNB with core biopsy where feasible; biopsy is generally required before neoadjuvant or palliative therapy, whereas patients with clearly resectable disease may proceed to surgery without preoperative biopsy. Measure baseline CA 19-9 (interpret after biliary decompression / with normal bilirubin; it can be falsely low in Lewis-antigen-negative patients and falsely elevated with cholestasis). Offer GERMLINE testing for inherited susceptibility to EVERY patient with confirmed pancreatic adenocarcinoma regardless of family history, using a comprehensive multi-gene panel; separately, for patients with locally advanced or metastatic disease who are candidates for anti-cancer therapy, obtain TUMOR/SOMATIC molecular profiling (broad NGS; RNA-based assays improve fusion detection; cell-free/liquid biopsy if tissue is insufficient). Assign resectability, resectable, borderline resectable, locally advanced (unresectable), or metastatic, by multidisciplinary review, ideally at a high-volume center. RESECTABLE disease: upfront resection (pancreaticoduodenectomy for head/uncinate; distal pancreatectomy ± splenectomy for body/tail) followed by adjuvant therapy, or neoadjuvant therapy in selected high-risk patients; preferred adjuvant regimen is modified FOLFIRINOX in patients able to tolerate it, ideally begun within 12 weeks of surgery, for a total of ~6 months of perioperative systemic therapy. BORDERLINE RESECTABLE and LOCALLY ADVANCED disease: neoadjuvant systemic therapy, with or without subsequent chemoradiation/RT, then restaging and reassessment for resection. METASTATIC disease by performance status: good PS (ECOG 0-1, adequate organ function), first-line FOLFIRINOX/modified FOLFIRINOX, NALIRIFOX, or gemcitabine + nab-paclitaxel, with clinical trial preferred; after ~4-6 months without progression, transition to a maintenance strategy (including olaparib for germline BRCA1/2 carriers who have NOT progressed on first-line platinum-based therapy), a clinical trial, or a treatment break; intermediate PS (ECOG 2), clinical trial preferred, otherwise less-intensive first-line therapy (preferred gemcitabine + nab-paclitaxel [category 1], or single-agent gemcitabine or capecitabine, or FOLFOX/FOLFIRI/CapeOx), since the intensive triplets (FOLFIRINOX/modified FOLFIRINOX and NALIRIFOX) are limited to ECOG 0-1 and are generally not appropriate at ECOG 2, with the same maintenance-versus-progression reassessment applied after ~4-6 months without progression; poor PS (ECOG 3-4), palliative/best supportive care, with single-agent chemotherapy or a biomarker-directed agent only if clinically appropriate. Treat by molecular target wherever an actionable alteration is present, across lines: pembrolizumab (or dostarlimab) for MSI-H/dMMR and pembrolizumab for TMB-H, olaparib for germline BRCA1/2 (maintenance after platinum), larotrectinib/entrectinib for NTRK fusions, selpercatinib for RET fusions, dabrafenib + trametinib for BRAF V600E, erdafitinib for FGFR2 alterations, fam-trastuzumab deruxtecan for HER2-positive disease, zenocutuzumab for NRG1 fusion-positive disease, and (in recent NCCN versions) KRAS G12C inhibitors (adagrasib/sotorasib), several of these after prior systemic therapy. Integrate supportive care throughout: relieve obstructive jaundice with an endoscopic self-expanding metal stent in non-operative or neoadjuvant pathways, but do NOT perform routine preoperative biliary drainage before planned upfront resection; treat exocrine insufficiency with pancreatic enzyme replacement titrated to symptoms (common starting dose ~40,000-50,000 units of lipase with main meals, roughly half with snacks, increased as needed); manage refractory pain with celiac plexus neurolysis or palliative RT; relieve gastric outlet obstruction with surgical or EUS-guided gastrojejunostomy (or enteral stenting); and provide VTE prophylaxis, nutritional support, and early palliative care.

In a patient with proctitis symptoms and receptive anal sex history, test with rectal NAAT for gonorrhea and chlamydia (LGV testing if positive), syphilis serology, HSV PCR from ulcers, stool studies for enteric pathogens in proctocolitis, plus HIV and hepatitis testing. Give empiric ceftriaxone 500 mg IM once plus doxycycline 100 mg BID for 7 days presumptively while awaiting results, started whenever the presentation is consistent with acute proctitis (anorectal exudate on exam or polymorphonuclear leukocytes on a Gram-stained anorectal smear, or, if anoscopy/Gram stain is unavailable, in a patient reporting receptive anal exposure) rather than only once symptoms are "moderate-severe", then tailor therapy by pathogen (extend doxycycline to the full 21-day course presumptively for LGV when bloody discharge, perianal or mucosal ulcers, or tenesmus accompany a positive rectal chlamydia NAAT, rather than waiting for confirmed LGV; penicillin G benzathine for syphilis, antivirals for HSV). Include partner notification and treatment, sexual health counseling, and repeat STI testing at 3 months.